microbial degradation of quaternary ammonium alcohols
TRANSCRIPT
Research Collection
Doctoral Thesis
Microbial degradation of quaternary ammonium alcoholshydrolysis products of esterquat surfactants used as fabricsofteners
Author(s): Käch, Andres
Publication Date: 2002
Permanent Link: https://doi.org/10.3929/ethz-a-004423464
Rights / License: In Copyright - Non-Commercial Use Permitted
This page was generated automatically upon download from the ETH Zurich Research Collection. For moreinformation please consult the Terms of use.
ETH Library
DISS. ETH NO. 14757
Microbial degradation of quaternary ammonium alcohols
hydrolysis products of esterquat surfactants used as fabric softeners
A dissertation submitted to the
SWISS FEDERAL INSTITUTE OF TECHNOLOGY ZURICH
for the degree of
Doctor ofNatural Sciences
presented by
ANDRESKAECH
Dip!. Natw. ETH
born 10.03.1969
Citizen ofEmmen, Lucerne
accepted on the recommendation of
Prof. Dr. A. J. B. Zehnder, examiner
Dr. N. Rehman, co-examiner
PD Dr. T. Egli, co-examiner
Zurich, 2002
.(1' ite Leer /I
, lank leaf..._._-------'
Dank
Allen voran mochte ich PD Dr. Thomas Egli fUr die Betreuung und Untersttitzung wahrend
meiner Dissertation sowie fUr seinen immerwahrenden Optimismus danken.
Dank gebtihrt auch:
Prof. Alexander Zehnder fUr die Ubemahme des Referates und wertvolle Diskussionen.
Steve G. Hales fUr die Initiation dieser Arbeit, Naheed Rehman fUr die FortfUhrung der
Betreuung seitens Unilever und die Ubemahme des Korreferates sowie Unilever (SEAC
Applied Science & Technology, Unilever Colworth, UK) fUr die Finanzierung dieser Arbeit.
Martina Hofer, die mit ihrer Diplomarbeit einen wesentlichen Beitrag an diese Arbeit geleistet
und mich fUr meine letzte Laborzeit nochmals so richtig motiviert hat.
Nathalie Vallotton fUr ihren grossen Beitrag an die Charakterisierung der isolierten Bakterien
wahrend ihres Praktikums.
Henri Lambert fUr die wertvolle Arbeit wahrend seiner Diplomarbeit.
Dem Laborantenteam, namentlich Thomi, Hansueli, Christoph, Teresa, Karin, Andy und
Bettina fUr ihre Hilfsbereitschaft, technische Beratung und Gesellschaft.
Der Gruppe Egli fUr das angenehme Arbeitsklima und hilfreiche Diskussionen und
insbesondere allen MitbastlerInnen im F 51, welche wesentlich zur Verarbeitung meiner
Hohen und Tiefen beigetragen haben, speziell Lukas, meinem langjahrigen Labomachbar.
Dani Rentsch fUr die Messung und Interpretation unzahliger NMR Spektren und fUr wertvolle
sowie kritische Diskussionen.
Wemer Angst fUr die Beratung in organisch chemischen Fragen.
Dem ganzen Prozess MIX fUr die gute Atmosphare.
Den Bibliothekarinnen fUr ihre stete Hilfsbereitschaft und prompten Lieferungen der
manchmal nicht leicht zu bekommenden Literatur.
Meinen Eltem, welche mich bisher in allem, was ich untemommen habe, jederzeit untersttitzt
haben.
"'- ._------,
ISeite Leer /Blank leaf
L-__. _----J
Publications
The following part of this thesis has been published:
Kaech, A. & Egli, T. (2001). Isolation and characterization of a Pseudomonas putida strain
able to grow with trimethyl-l,2-dihydroxy-propyl-ammonium as sole source of carbon,
energy and nitrogen. Syst Appl Microbiol24, 252-261. (Chapter 2)
The following publications are in preparation:
Kaech, A., Vallotton, N. & Egli, T. Isolation and characterisation of microorganisms able to
grow with quaternary ammonium alcohols as sole source of carbon, energy and nitrogen.
(Chapter 3)
Kaech, A., Hofer, M., Rentsch, D. & Egli, T. Microbial oxidation of methyl-triethanol
ammonium. (Chapter 4)
ISeite Leer /Blank leaf
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Table of contents
Table of contents
Abbreviations 9
Summary 11
Zusammenfassung 13
1. General introduction 17
2. Isolation and characterisation of a Pseudomonas putida strain able to grow
with 2,3-dihydroxypropyl-trimethyl-ammonium as sole source of carbon,
energy and nitrogen 23
3. Isolation and characterisation of bacteria able to grow with quaternary
ammonium alcohols as sole source of carbon, energy and nitrogen 43
4. Microbial oxidation of methyl-triethanol-ammonium 65
5. Microbial degradation of 2,3-dihydroxypropyl-trimethyl-ammonium 89
6. Concluding remarks 111
References 117
7
eite Leer /I Blank leaf!1, --'
Abbreviations
BSA
BTF
CE
CFE
CHD
D
DM
DOC
DON
dpm
Glycidol
INT
KEGG
MM
OD
OECD
PB
PF
PHA
PMS
QAA
SF
SM
TM
TMA
TSA
TSB
TSP
Bovine serum albumin
Benzotrifluorid
Crude extract
Cell-free extract
Choline dehydrogenase
Dilution rate
Dimethyl-diethanol-ammoniurn
Dissolved organic carbon
Dissolved organic nitrogen
Decays per minute
(±)-Oxiran-2-methanol
Iodonitrotetrazolium chloride
Kyoto Encyclopedia of Genes and Genomes
Methyl-triethanol-ammonium
Optical density
Organization for Economic Cooperation and Development
Phosphate buffer (50 mM)
Particulate fraction of CFE
Polyhydroxyalkanoate
Phenazine methosulfate
Quaternary ammonium alcohol
Soluble fraction of CFE
Synthetic medium
(±)-2,3-Dihydroxypropyl-trimethyl-ammonium
Trimethylamine
Tryptic soy agar
Tryptic soy broth
Sodium 3-trimethylsilyltetradeutero-propionate
9
Abbreviations
Seite Leer /1Blank I af
1..-- . __..__
Summary
Summary
The quaternary ammonium alcohols (QAAs) 2,3-dihydroxypropyl-trimethyl-ammonium
(TM), dimethyl-diethanol-ammonium (DM) and methyl-triethanol-ammonium (MM) are the
mainly used head groups in esterquat surfactants, which are widely applied as softeners in
fabric care. Four bacterial strains able to grow with these QAAs as the sole source of carbon,
energy and nitrogen were isolated. One strain was isolated with each, TM and MM, referred
to as strain TM 1 and MM 1, respectively. Two strains were isolated with DM, designated as
strain DM 1 and DM 2. Phylogenetic identification was performed and the morphology of
cells and colonies, the nutritional and biochemical properties as well as the growth
characteristics were investigated. Phylogenetic analysis (16S-rDNA) revealed for strain TM 1
identities to Pseudomonas putida DSM 291T of 99.9 % (closest relationship). For the strains
DM 1 and DM 2 the analysis provided closest relationship to Zoogloea ramigera Itzigsohn
1868AL of 97 and 98 % identities, respectively. Closest related strain for isolate MM 1 was
found to be Rhodobacter sphaeroides with identities of 94 % only. However, anaerobic
growth in the light and the formation of pigments characteristic for all representatives of the
genus Rhodobacter was not observed for strain MM 1. Therefore, strain MM 1 must be
considered to be a member of a new genus.
Out of the four investigated strains, only strain DM 2 isolated with DM was able to grow with
another QAA (TM) than that used for its isolation. No consortia of microorganisms were
required for complete degradation of the three QAAs. All strains were able to grow with the
naturally related compound choline, but none of the choline degrading bacterial reference
strains tested was able to grow with any of the QAAs. Therefore, the ability to degrade
choline does not go along with the competence to catabolise the three QAAs. The isolated
strains belonged to different genera and hence, the degradation of QAAs is not a trait of a
single microbial genus.
The primary catabolism of the QAAs was investigated and the metabolites were identified by
NMR spectroscopy. The initial enzymatic attack on MM in isolate MM 1 was mediated by a
membrane-associated, constitutively expressed oxidoreductase, oxidising ethanol groups first
to the corresponding aldehydes and then to the carboxylic acids. However, as soon as one
ethanol group of MM was oxidised to the aldehyde, a cyclisation occurred intramolecularly
with a second ethanol group forming a cyclic hemiacetal. No further oxidation of the cyclic
11
Summary
hemiacetal was observed. Only the remaining third ethanol group of MM was oxidised to the
aldehyde and to the carboxylic acid. The cyclic hemiacetal products with the third ethanol
group oxidised twice to the carboxylic acid appeared to be dead-end metabolites, since in
batch cultures of strain MM I considerable amounts of this compound were released to the
medium and remained untouched. Hence, these metabolites may accumulate in the
environment. The oxidation of the ethanol groups of DM and choline proceeded under the
same assay conditions as for the oxidation of MM, providing one cyclic hemiacetal product
from DM and betainealdehyde and betaine from choline. This suggests that the same enzyme
was responsible for the oxidation of MM, DM and choline. It might be a choline
oxidoreductase with extended substrate specificity.
In contrast, the initial attack on TM in isolate TM 1 was catalised by an inducible, membrane
associated lyase removing trimethylamine from the molecule. In the cell-free extracts of this
strain, enzymatic transformation of neither DM, MM, betaine nor carnitine was observed.
Only choline was transformed, however, it did not undergo a C-N fission as detected for TM,
but was oxidised to betainealdehyde and betaine, as found in the MM-growing strain MM 1.
Therefore, in strain TM 1 the mechanisms for the initial attack on TM and choline,
respectively, are different. When grown with TM, the fission of the C-N bond of TM and the
release of trimethylamine were also found in the cell-free extracts of the isolate DM 2 able to
grow with DM and TM, although this strain belongs to a different genus.
In the cellular fractions of the DM-growing strains DM 1 and DM 2 no DM-consuming
activity was found and therefore, no enzymatic investigation of the catabolism of DM was
possible.
Based on these findings, no general strategy in microorganisms can be proposed for the
degradation of QAAs despite the similarity of these compounds in their chemical structure.
Also, no general relationship seems to exist between the degradation of the QAAs and the
naturally related compound choline. The ability to degrade the QAAs appears to be a specific
capability of specialised microorganisms, and individual degradation mechanisms seem to be
followed for each of the three QAAs.
12
Zusammenfassung
Zusammenfassung
Die quatemaren Ammoniumalkohole (QAAs) 2,3-Dihydroxypropyl-trimethyl-ammonium
(TM), Dimethyl-diethanol-ammonium (DM) und Methyl-triethanol-ammonium (MM) sind
die meist verwendeten Kopfgruppen in Esterquat-Tensiden, welche als Weichmacher in
grossen Mengen in der Textilreinigung eingesetzt werden und weit verbreitet sind. Vier
bakterielle Stamme, welche mit den entsprechenden QAAs als alleinige Kohlenstoff-,
Energie- und Stickstoffquelle wachsen konnen, wurden isoliert. Je ein Stamm wurde mit TM
und MM isoliert, bezeichnet als Stamm TM 1 beziehungsweise MM 1. Mit DM wurden zwei
Stamme isoliert, Stamm DM 1 und DM 2. Die Isolate wurden phylogenetisch identifiziert,
und die Morphologie der Zellen und Kolonien, das Niihrstoffspektrum, diverse biochemische
Eigenschaften sowie die Wachstums-Charakteristik wurden eingehend untersucht. Die
phylogenetische Analyse (16S-rDNA) ergab fur den Stamm TM 1 eine Obereinstimmung von
99.9 % mit Pseudomonas putida DSM 291 T (nachste Verwandtschaft). Fiir die Stamme DM 1
und DM 2 lieferte die Analyse eine nachste Verwandtschaft zu Zoogloea ramigera Itzigsohn
1868AL mit einer Obereinstimmung von 97 beziehungsweise 98 %. Der nachst verwandte
Stamm von Isolat MM 1 war Rhodobacter sphaeroides mit einer Obereinstimmung von nur
94 %. Da weder anaerobes Wachstum mit Licht noch die Bildung von Pigmenten fur Stamm
MM 1 beobachtet wurde, beides Eigenschaften aller Vertreter der Gattung Rhodobacter, ist
Isolat MM 1 hochst wahrscheinlich ein Vertreter einer neuen Gattung. Von den vier isolierten
Bakterienstammen war nur Isolat DM 2, isoliert mit DM, dazu befahigt, mit einem der
anderen QAAs (TM) zu wachsen. Es waren keine Konsortien von verschiedenen Bakterien
fur den Abbau der drei QAAs notwendig. Alle isolierten Bakterien konnten mit der
natiirlichen, strukturell verwandten Substanz Cholin wachsen. Jedoch konnte keiner der
getesteten cholinabbauenden bakteriellen Referenzorganismen mit einem der QAAs wachsen.
Daher impliziert die Fahigkeit Cholin abzubauen nicht notwendigerweise die Fahigkeit, auch
die QAAs zu verwerten. Da die isolierten Sllimme zu verschiedenen Gattungen gehoren, ist
der Abbau der QAAs nicht die Vorherrschaft einer einzelnen bakteriellen Gattung.
Der anfangliche Abbaustoffwechsel der QAAs wurde untersucht und die
Umwandlungsprodukte wurden mit Hilfe der NMR-Spektroskopie identifiziert. Der initiale
enzymatische Abbauschritt von MM in Isolat MM 1 wurde durch eine membrangebundene,
konstitutiv exprimierte Oxidoreduktase katalysiert, welche Ethanolgruppen zuerst zum
13
Zusammenfassung
entsprechenden Aldehyd und dann zur Carboxysaure oxidiert. Sobald jedoch eine
Ethanolgruppe von MM zum Aldehyd oxidiert war, fand eine intramolekulare Zyklisierung
mit einer zweiten Ethanolgruppe statt, welche zu einem zyklischen Hemiacetal fUhrte. Eine
weitere Oxidation des zyklischen Hemiacetals wurde nicht beobachtet. Nur die
tibrigbleibende dritte Ethanolgruppe von MM wurde zum Aldehyd und zur Saure oxidiert.
Die zyklischen Hemiacetal-Produkte, deren dritte Ethanolgruppe jeweils zur Carboxysaure
oxidiert war, scheinen Sackgasse-Abbauprodukte zu sein, da in Batchkulturen des Stammes
MM 1 erhebliche Mengen dieser Zwischenprodukte ausgeschieden und nicht weiterverwertet
wurden. Diese Zwischenprodukte konnten sich daher hypothetisch in der Umwelt anreichem.
Die Oxidation der Ethanolgruppen von DM und Cholin erfolgte unter den gleichen
experimentellen Bedingungen wie die Oxidation von MM, wobei hauptsachlich ein
zyklisches Hemiacetal-Produkt aus DM und Betainaldehyd und Betain aus Choline gebildet
wurde. Daher war das aktive Enzym vermutlich dasselbe fUr MM, DM und Cholin. Es konnte
sich urn eine Cholin Oxidoreduktase mit erweitertem Substratspektrum handeln.
Im Gegensatz dazu beruhte der initiale Abbauschritt von TM in Stamm TM 1 auf einem
anderen Mechanismus. Es wurde eine induzierbare, membrangebundene Lyaseaktivitat
gefunden, welche Trimethylamin von TM abspaltet. In den zellfreien Extrakten dieses
Stammes wurden weder DM, MM, Betain noch Camitin umgewandelt. Nur Cholin wurde
abgebaut, jedoch nicht durch eine Spaltung der C-N Bindung wie bei TM, sondem durch eine
Oxidation zu Betainaldehyd und Betain, wie dies auch im MM-wachsenden Stamm MM 1
beobachtet wurde. Damit sind die Mechanismen des initialen Abbauschrittes von TM
beziehungsweise Cholin in Stamm TM 1 ganzlich unterschiedlich. Die Spaltung der C-N
Bindung von TM und die Freisetzung von Trimethylamin wurde auch im zellfreien Extrakt
des mit TM gewachsenen Stammes DM 2 gefunden, welcher mit DM und TM wachsen kann,
obwohl dieser Stamm zu einer anderen Gattung gehort.
In den Zellaufschltissen der mit DM wachsenden Stamme DM 1 und DM 2 wurde keine
Enzymaktivitat fUr DM gefunden und deshalb konnten keine enzymatischen Untersuchen
beztiglich des Abbaustoffwechsels von DM durchgeftihrt werden.
Aufgrund dieser Resultate kann trotz der ahnlichen chemischen Struktur dieser Substanzen
keine generelle Strategie in Mikroorganismen fUr den Abbau der QAAs vorgeschlagen
werden. Zudem scheint keine allgemeine Beziehung zwischen dem Abbau der QAAs und der
14
Zusammenfassung
nattirlich verwandten Substanz Cholin zu bestehen. Die Fahigkeit, QAAs abzubauen, ist
offenbar eine Fahigkeit von spezialisierten Mikroorganismen, und die gewahlten
Abbaumechanismen scheinen von der spezifischen Struktur der QAAs abhangig zu sein.
15
Seite Leer /1Blank leaf I
I
General introduction
1. General introduction
In the 1950s and 1960s, a new group of surfactants was successfully introduced in the
detergent and cleaning agent sector. These were called softeners because they eliminated the
harsh feel produced by modern laundry processes. Softeners, however, do not only soften
tissues. They have a number of other useful properties such as preventing accumulation of
electrostatic charge, improving the suppleness of the fibers, making ironing easier, shorten
drying time, and they impart a pleasant fragrance to laundry. Because of the many
advantageous properties and effects, they very soon became popular and, hence, they present
now a very important group of products (Puchta et aI., 1993).
Not many different chemicals are used as softeners. For more than three decades (up to 1990)
almost exclusively one raw material was applied. This material was called DTDMAC
ditallow-dimethyl-ammonium chloride. It is a quaternary ammonium compound with two
methyl groups and two alkyl groups derived from tallow connected to the nitrogen atom
(Figure 1. la). Commercial DTDMAC typically comprised C18 (65 %), C16 (30 %) and C14
(5 %) alkyl chains (Sullivan, 1982). In the 1990s DTDMAC was given an environmentally
hazardous classification as a result from laboratory studies suggesting that the Predicted
Environmental Concentration (PEC) of DTDMAC would exceed its No-Effect Concentration
(NEC), particularly in surface waters charged heavily with treated wastewater. Although a
detailed study published later by the European Center for Ecotoxicology and Toxicology of
Chemicals (ECETOC) came to the conclusion that the use of DTDMAC is safe and its
environmental concentrations do not pose a risk to aquatic and terrestrial ecosystems, the
industry at that time reacted rapidly and a new group of softeners with improved ecological
properties was introduced in several countries (Berenbold, 1990; Krueger et aI., 1998; Puchta
et al., 1993). The difference to the DTDMAC based softeners consisted in the insertion of
esters groups into the alkyl chains, and due to this chemical structure they were named
esterquat surfactants. Three similar esterquat surfactants were developed and are still today
the mainly used softeners (Figure 1.1b). The annual production exceeded in the 1990s
probably 100'000 tons worldwide, with basically over 99 % of the material being used in
fabric care (Krueger et aI., 1998). These esterquat softeners have proven to fulfill all the
properties required for good washing. Based on standardised test procedures, they were
expected to be readily and ultimately biodegradable and to cause no risks for human health
17
Chapter 1
(a) DTDMAC*
(b) Esterquat surfactants
oOAR
,,( I/~OyR
oTM-esterquat
tdrOIYSiS
~ ........ fatty acid
DM-esterquat
tdrOIYSiS
~ ........ fattYacid
MM-esterquat
tdrOIYSiS
~ ........ fatty acid
(c) Quaternary ammonium alcohols (QAAs)
/ OH
"N~OH/
TM
HO /"--.../OH~N+
/~OH
MM
(d) Choline, structurally related compound of natural origin
Rand *: Fatty acid carbon chain derived from tallow, mainly C16 and C1S.
Figure 1.1. Structure of Ca) DTDMAC, Cb) the esterquat surfactants used as fabric softeners, Cc) their
hydrolysis products, Cd) and the structurally related compound choline.
(Giolando et al., 1995; Krueger et al., 1998; Matthijs et al., 1995; Puchta et al., 1993; Waters
et aI., 1991; Waters et al., 2000).
The ester bonds, characteristic for these new softeners, serve as potential breaking points of
the molecule and, therefore, contribute to the good environmental behaviour of these
compounds (Figure 1.1b) (Puchta et al., 1993). They hydrolise rapidly, abiotically and/or
biocatalysed, when reaching surface water or sewage treatment plants. The resulting products
are the fatty acids and the corresponding quaternary ammonium alcohols (QAAs)
2,3-dihydroxypropyl-trimethyl-ammonium (TM), dimethyl-diethanol-ammonium (DM) and
methyl-triethanol-ammonium (MM) as displayed in Figure 1.1c. The fatty acids are common
compounds in the environment and hence are degraded readily and completely by many
different microorganisms by ~-oxidation. The QAAs were investigated as well by applying
18
General introduction
the same standardised test procedures as for the parent compounds and were also expected to
be readily and ultimately biodegradable in the environment and to have no harmful effects on
human health (Giolando et al., 1995; Hellberg et al., 2000; Krueger et al., 1998; Matthijs et
al., 1995; Puchta et al., 1993; Simms et al., 1992; Waters et al., 2000). However, it is still
possible that one of the QAAs might accumulate in the environment due to high amounts used
of the esterquat surfactants.
Surprisingly, the three structurally similar QAAs showed very different degradation patterns
and degradation rates in OECD die-away tests (OECD, 1981) mimicking complex
environmental systems, although all three QAAs were assessed as readily and ultimately
biodegradable (Hales, 1998). Considering the similarity of the three QAAs and their similarity
to the naturally related compound choline (Figure 1.1d), one could suggest that all of the three
QAAs should degrade in a similar way and with a rate comparable to that of choline.
However, this was not the case. The question arised as to why the three QAAs behave
differently concerning their biodegradation and what mechanisms might be responsible for
their degradation. Was it due to the limited distribution and occurrence of QAA degrading
microbes or to their QAA degrading properties? Furthermore, since choline is a widespread
compound among microorganisms as well as in higher organisms (Kortstee, 1970) the
question came up whether the mechanisms of QAA breakdown are related to those involved
in the degradation of choline.
Up to now, no microorganisms degrading the three QAAs TM, DM and MM have been
isolated and consequently the catabolic pathways are not elucidated yet. Knowledge of these
mechanisms is very important considering the high amounts of these compounds used and the
increasing market of a whole group of similar compounds. Moreover, a better understanding
could affect the selection of more favorable and the design of new similar compounds, which
are not only used as raw material for softeners but also for cosmetics, drugs and other
chemicals used in biological applications (Krueger et al., 1998; Vievsky, 1997).
Several studies report on the microbial degradation of choline. The general degradation
pathway (Figure 1.2a) present in many and widely different microbial strains proceeds by
successive oxidation to betainealdehyde and betaine, followed by progressive demethylation
to dimethylglycine, sarcosine and finally the amino acid glycine (Kortstee, 1970; Shieh, 1964;
19
Chapter 1
KEGG Kyoto Encyclopedia of Genes and Genomes, www.genome.ad.jp). Several bacterial
enzymes mediating initial choline oxidation have been described in the literature (Bater &
Venables, 1977; Haubrich & Gerber, 1981; Ikuta et al., 1977; Kiene, 1998; Nagasawa et aI.,
1975; Nagasawa et aI., 1976; Ohta-Fukuyama et aI., 1980; Rosenstein et al., 1999; Yamada et
aI., 1979). Additionally to initial choline oxidation, a fission of the C-N bond, producing
trimethylamine, was observed in Proteus vulgaris (Seim et aI., 1982b) and Shigella
alkalescens (Wood & Keeping, 1944). In both cases, trimethylamine was released into the
culture medium.
In contrast, only a limited number of studies on the isolation of bacteria able to degrade
human-made quaternary ammonium compounds have been reported so far (Dean-Raymond &
Alexander, 1977; Nishihara et aI., 2000; Van Ginkel et aI., 1992). Microorganisms were
isolated from activated sewage sludge or soil using the quaternary ammonium surfactants
decyl-trimethyl-ammonium, hexadecyl-trimethyl-ammonium and didecyl-dimethyl
ammonium as sole source of carbon. However, the breakdown of these compounds by pure
cultures was always incomplete and required a consortium of at least two different
microorganisms for complete degradation. Based on these studies and other investigations on
the biodegradation of a series of anionic, non-ionic, amphoteric and cationic surfactants Van
Ginkel (1996) concluded that complete degradation of most surfactants has to be achieved by
consortia of microorganisms. This author suggested that only a few surfactants, i. e. alkane
sulphonates, alkyl sulphates and alkylamines are completely degraded by pure microbial
cultures. Three possible degradation mechanisms of alkyl-trimethyl-ammonium compounds
have been postulated by Van Ginkel (1995) (Figure 1.2b):
(1) Oxidation at the free end of the alkyl chain followed by stepwise degradation via
~-oxidation (fatty acid metabolism) resulting in betaine, which is oxidatively
demethylated and finally results in the amino acid glycine (Figure 1.2a).
(2) Progressive demethylation of the nitrogen centre, then the splitting off of the nitrogen
from the alkyl chain in the form of ammonia. The alkyl chain again is degraded by
~-oxidation.
(3) Oxidative cleavage of the N-Calkyl bond providing trimethylamine and the aldehyde of the
alkyl chain. Whereas the alkyl chain again undergoes ~-oxidation, trimethylamine is
degraded by methylotrophic organisms.
20
General introduction
(a)
/ "U/ OH
"N+ N+ II "N~/~OH • / 0 •
III III / 0
Choline Betainealdehyde Betaine
IvlOH VI I OH V I OH
H2N~O .. HN~ .. /N~OVII 0
Glycine Sarcosine Dimethylglycine
(b)
o
H~~Betaine
Figure 1.2. (a) General schematic degradation pathway of choline with responsible enzymes, adapted
from the Kyoto Encyclopedia of Genes and Genomes (www.genome.ad.jp): I) Choline dehydrogenase
(EC 1.1.99.1), II) betainealdehyde dehydrogenase (EC 1.2.1.8), Ill) choline oxidase (EC 1.1.3.17), IV)
betaine homocysteine S-methyltransferase (EC 2.1.1.5), V) dimethylglycine dehydrogenase
(EC 1.5.99.2), VI) sarcosine dehydrogenase (EC 1.5.99.1), VII) sarcosine oxidase (EC 1.5.3.1).
(b) Schematic degradation mechanisms of alkyl-trimethyl-ammonium compounds as proposed by Van
Ginkel (1995); numbers correspond to the explanations in the text.
21
Chapter 1
Van Ginkel (1996) proposed that the initial fission of the N-Calkyl bond represents a general
strategy of microorganisms to gain access to the alkyl chains of quaternary ammonium
compounds.
The primary goal of this thesis was the elucidation of the strategies of microorganisms to
degrade the quaternary ammonium alcohols TM, DM and MM. To approach this goal, the
isolation of competent microorganisms degrading the QAAs is an indispensable requirement,
since these bacteria are the basis for further investigations at the enzyme level. With respect to
the aim set the following questions were considered to be of primary importance:
• Can pure microbial cultures or defined consortia be enriched and isolated that are able to
grow with the three QAAs TM, DM and MM as sole sources of carbon, energy and
nitrogen and are capable to degrade them to completion?
• Are the strains isolated on the different QAAs totally different, similar, or even the same?
• How specific is the degradation of the different QAAs. Are the different isolated strains
able to degrade the two other QAAs, not used for their isolation, too?
• Are the strains able to grow with the naturally related compound choline?
• What kinds of enzymes are responsible for the first step in the catabolism of the individual
QAAs?
• Are the enzymes of the different microorganisms, attacking the different QAAs, the same
or different?
• How specific are the enzymes with respect to structurally similar compounds?
• Are the enzymes similar or identical to those responsible for the degradation of choline?
22
Pseudomonas putida TM 1
2. Isolation and characterisation of a Pseudomonas putida strain able to
grow with 2,3-dihydroxypropyl-trimethyl-ammonium as sole source of
carbon, energy and nitrogen
ABSTRACT
2,3-dihydroxypropyl-trimethyl-ammonium (TM) originates from the hydrolysis of the parent
esterquat surfactant, which is widely used as softener in fabric care. Based on test procedures
mimicking complex biological systems, TM is supposed to degrade completely when
reaching the environment. However, no organisms able to degrade TM were isolated nor has
the degradation pathway been elucidated so far. We isolated a Gram-negative rod able to
grow with TM as sole source of carbon, energy and nitrogen. The strain reached a maximum
specific growth rate of 0.4 h- l when growing with TM as the sole source of carbon, energy
and nitrogen. TM was degraded to completion and surplus nitrogen was excreted as
ammonium into the growth medium. A high percentage of the carbon in TM (68 % in
continuous culture and 60 % in batch culture) was combusted to C02 resulting in a low yield
of 0.54 mg cell dry weight per mg carbon during continuous cultivation and 0.73 mg cell dry
weight per mg carbon in batch cultures. Choline, a natural structurally related compound,
served as a growth substrate, whereas a couple of similar other quaternary ammonium
alcohols also used in softeners did not. The isolated bacterium was identified by 16S-rDNA
sequencing as a strain of Pseudomonas putida with a difference of only one base pair to
P. putida DSM 291T• Despite their high similarity, the reference strain P. putida DSM 291T
was not able to grow with TM and the two strains differed even in shape when growing on the
same medium. This is the first microbial isolate able to degrade a quaternary ammonium
softener head group to completion. Previously described strains growing on quaternary
ammonium surfactants (decyl-trimethyl-ammonium, hexadecyl-trimethyl-ammonium and
didecyl-dimethyl-ammonium) either excreted metabolites or a consortium of bacteria was
required for complete degradation.
23
Chapter 2
INTRODUCTION
Today, mainly three structurally similar "esterquats", belonging to the group of cationic
surfactants, are used as laundry softeners in detergents (Figure 2.1). The worldwide annual
production of these surfactants probably exceeds 100'000 tons, with basically over 99 % of the
material being used in fabric care (Krueger et aI., 1998). The three "esterquats" appear to
hydrolyse rapidly (abiotically and/or biocatalysed) when reaching surface water or sewage
treatment plants, with the corresponding fatty acids and three quaternary ammonium alcohols
(QAAs) being the products (Giolando et aI., 1995; Hellberg et aI., 2000; Krueger et aI., 1998;
Puchta et aI., 1993; Simms et aI., 1992; Waters et al., 1991). The resulting QAAs are
trimethyl-2,3-dihydroxy-propyl-ammonium (TM), dimethyl-diethanol-ammonium (DM) and
methyl-triethanol-ammonium (MM) (Figure 2.1). The biodegradation properties of the parent
esterquat surfactant and the QAAs have been investigated in a variety of test procedures
Esterquat surfactants
oII
~c~~--O-C-R
I. ~H3C~I-C~-c~-0-C-- R
c~---c~-~ OH
MM-esterquat
lhydrolysis
f" fatty acids
lhydrolysis
f''''- fatty acids
Quaternary ammonium alcohols (QAAs)
o11
CH3 ~C~O--C-R
I. I ~~C~N-C~-CH-O-C~-R
ICH3
TM-esterquat
lhydrolysis
f'''- fatty acids
R: Fatty acid carbon chain derived from tallow, mainly C16 and C18
o11
~C~C~-O-C-R
I. ~H3C--~ ~C~-c~ -O-C- R
ICH3
DM-esterquat
CH3 ~C~OH
I. 1H3C - ~ ~-C~ ~- CH-~ OH
CH3 TM
H2T ~~- C~ -~~ OH
H C-~N·--CH -~C~ -OH3 I 2 • '2
CH3 OM
~ C--CH --OH• '2
1
2
H3c-i-c~-C~-OH
C~-~-OH MM
Choline, naturally related compound
CH3
I.H3C--i-- . C~--CH2-0H
C~
Figure 2.1. Structures of the three main commercially used esterquat surfactants, their hydrolysis
products and structure of the naturally related compound choline.
24
Pseudomonas putida TM 1
mimicking degradation in complex systems, for instance activated sewage treatment sludge.
Based on the performed tests, the parent esterquats and the QAAs are expected to be readily
and ultimately biodegradable (Giolando et aI., 1995; Krueger et aI., 1998; Matthijs et aI.,
1995; Puchta et aI., 1993; Simms et al., 1992; Waters et aI., 1991; Waters et aI., 2000).
The similarity of the QAAs and their structural relationship to the naturally occurring
compound choline (Figure 2.1) suggests that all three QAAs should degrade in a similar way
and with a similar rate, comparable to that of choline. However, despite their similarity they
show different degradation patterns as well as different degradation rates in GECD die-away
tests (Hales, 1998), and several choline-degraders, including one reference strain from DSM,
were not able to degrade any of the QAAs (own results). Also from the simple and choline
like structure of the molecule one would expect that many different microorganisms would be
able to utilise the compound as a single source of carbon, energy and nitrogen and that a
consortium would not be required for complete degradation. However, to date, no organisms
able to degrade these QAAs have been isolated and consequently the catabolic pathway for
neither of them has been elucidated yet. Considering the wide application of esterquat
surfactants as softeners in laundry detergents, we have therefore set out to enrich and isolate
microbes able to grow with these QAAs as their only source of carbon, energy and nitrogen.
Two different microorganisms able to grow with TM have been isolated and one strain,
designated TM 1, was selected for closer examination due to its high growth rate and
reproducible growth.
MATERIALS AND METHODS
Chemicals. The quaternary ammonium alcohols (QAA) (±)-2,3-dihydroxypropyl-trimethyl
ammonium (TM), dimethyl-diethanol-ammonium (DM) and methyl-triethanol-ammonium
(MM) were provided by Unilever (SEAC Safety and Environmental Assessment Center,
Unilever Research, Port Sunlight, UK) as the iodine salts. All other compounds were
purchased from Fluka, Buchs, Switzerland.
Isolation, growth and maintenance of organisms. For isolation of organisms (in batch and
continuous enrichment cultures), for growth tests and batch experiments, the following
25
Chapter 2
synthetic medium (SM) was used. It contained per litre of deionised water: MgS04'7HzO,
0.3 g; CaClz'2HzO, 0.02 g; NazHP04'2HzO, 2.05 g; KHZP04, 1.30 g; 1 ml of trace element
stock solution as described by Pfennig et al. (1981), but 3-times concentrated (containing per
litre: FeClz'4HzO, 4.5 g; MnClz'4HzO, 0.3 g; CoClz'6HzO, 0.36 g; ZnClz, 0.21 g;
CuClz'2HzO, 0.045 g; NazMo04'2HzO, 0.075 g; H3B03, 0.18 g; NiClz'6HzO, 0.075 g;
Na4EDTA'4HzO, 14.023 g); 1 ml of vitamin stock solution (which contained per litre:
pyridoxin'HCl, 100 mg; 50 mg of each, thiamine'HCI, riboflavin, nicotinic acid, D-Ca
pantothenic acid, p-amino benzoic acid, lipoic acid, nicotinamide, vitamin BIZ; biotin 20 mg
and folic acid 20 mg). The pH of the medium was always 7.0 except for experiments used for
the determination of pH optimum. For continuous cultivation the described medium was
slightly changed. Phosphate was exclusively added as KHZP04 (1.2 g rI) and the medium was
acidified with 0.2 ml r I of concentrated HZS04. The pH in the reactor was maintained at 7.0
by continuous addition of a NaOH/KOH (1 M/1 M) solution. Additionally, 0.1 ml r I of
antifoam (Silicon, emulsion 30 % in water, Fluka) was added to the medium to avoid
foaming. No vitamins were provided in continuous cultures and batches in bioreactors, since
isolate TM 1 was able to grow in the absence of vitamins.
The medium used for continuous cultivation was prepared as described above and it was
supplemented with the carbon source of choice after sterilisation by sterile filtration. For
sterilisation of batch media, MgS04'7HzO, CaClz·2HzO and trace elements were added to
nanopure water and autoclaved in Erlenmeyer flasks. Phosphate buffer, vitamins and the
carbon source of choice were added after sterilisation to the cooled-down medium by sterile
filtration. For substrate tests the complete medium was filtered sterile into heat-sterilised glass
test tubes. In each case, filtration was performed using sterile Millex-GP filters of 0.22 t-tm
pore size (Millipore, Volketswil, Switzerland).
If the growth substrate did not contain nitrogen, ammonia was added to the SM as N~CI,
0.8 g r I. For agar plates 1.5 % of highly purified agar (Biolife, Milano, Italy) was added to the
SM, both before sterilisation.
For experiments in complex liquid media, tryptic soy broth (TSB) (Biolife, Milano, Italy) was
used as a ten-fold dilution. Agar plates containing complex media of the appropriate tryptic
soy content were prepared by adding 13.5 g rI of highly purified agar to 4 g rI of tryptic soy
agar (TSA) (both Biolife, Milano, Italy).
26
Pseudomonas putida TM 1
Bacterial strains and storage. Pseudomonas putida DSM 291T was obtained from Deutsche
Sammlung von Mikroorganismen und Zellkulturen GmbH (DSMZ), Braunschweig,
Germany. For short-term storage all strains were plated on lO-fold diluted TSA or agar plates
containing SM and a selective carbon source. For long-term preservation all strains were
suspended in 30 % glycerol and stored at -80 QC. 16S-rDNA sequencing of strain TM 1 for
identification was performed by DSMZ.
Nutritional and biochemical properties. Utilisation of different carbon sources and selected
metabolic activities were investigated in various commercially available ready test kits such
as API 20 NE, API 50 CH (Biomerieux, Geneva, Switzerland) and Biolog GN plates (Biolog,
Hayward, California, USA) following the instructions of the manufacturer. For API 50 CH
and Biolog GN plates SM with N&CI (0.8 g r l) as nitrogen source was used. The utilisation
of selected carbon sources and combined carbon/nitrogen sources was tested in liquid culture
(5 ml in glass test tubes) at pH 7 using SM supplemented with the compound of choice (50
250 mg r l) by following turbidity. Controls without a carbon source were used. All tests were
inoculated with cells taken from colonies pregrown on tryptic soy agar. Cultures were
incubated at 30 QC in the dark and all tests were done in triplicates.
The Gram-reaction was tested with the classical staining procedure as described by Drews
(1968), the KOH method (Buck, 1982) and by testing for the presence of L-alanin
aminopeptidase (Bactident strips, Merck, Darmstadt, Germany).
Production of fluorescent pigments was tested by plating cells on agar containing an iron
deficient medium (Gould SI, containing per litre: Sucrose, 10 g; casamino acids, 5 g; glycerol
87 %, 10 ml; agar, 18 g; MgS04 1 M, 1.66 ml; CaCh 1 M, 0.18 ml).
SDS-PAGE of cellular proteins was performed as described by Laemmli (1970). Cells used
for SDS-PAGE were grown in batch cultures. Samples were taken in the exponential phase of
growth and harvested in Eppendorf tubes in a microcentrifuge (ALC, Milano, Italy). The
supernatant was discarded and the pellet stored at -20 QC not longer than one month prior to
SDS-PAGE.
Growth characteristics. Experiments for the determination of optimum growth conditions of
strain TM 1 were performed in batch cultures in Erlenmeyer flasks containing 100 ml of SM
27
Chapter 2
and TM (500 mg r l). Batch media were inoculated with cells pregrown in the same medium
taken from the exponential growth phase. Incubation was executed at different temperatures
(4, 12,20,25,28,30,32,37 QC) and pH (5.0; 6.0; 6.5; 7.0; 7.5; 8.0; 9.0). For batch growth at
different pH values the buffer stock solution was adjusted with either NaOH or HCI before
adding it to the medium. Growth was followed by measuring optical density at 546 nm
(ODS46) with a glass cuvette (1 cm) in a spectrophotometer (Kontron Uvikon, Zurich,
Switzerland). All experiments were performed at least in duplicate.
Determination of the maximum specific growth rate was performed for growth with TM at
optimum conditions following ODS46 under nutrient excess conditions in batch culture. For
this purpose, cells were transferred repeatedly into fresh medium before they had reached the
late exponential phase until they exhibited a stable JLmax'
Carbon and nitrogen balances were performed in batch and continuous culture both in a
bioreactor of 1 I working volume (Bioengineering AG, Wald, Switzerland) at 30 QC and pH
7.0. Chemostat cultures were run at a dilution rate (D) of 0.1 h- I and with a TM concentration
in the feed medium of 2 g r l. For DOC, TM, NHt, N02-, N03-, TMA, and Ntot analysis
samples were taken from the cultures and filtered (Millipore PVDF, 0.22 JLm membrane,
Millipore, Volketswil, Switzerland) using a vacuum pump and filtrates were stored at -20 QC
prior to analysis. Cell dry weight was measured by collecting cells from a known culture
volume on preweighed Nuclepore polycarbonate filters, 0.2 JLm pore size (Sterico AG,
Dietikon, Switzerland). Filters were washed once with distilled water and dried at 120 QC.
Dissolved organic carbon (DOC) was measured with a Tocor 2 Carbon Analyzer (Maihak,
Hamburg, Germany). ~+-N was determined by the indophenol method as described by
Scheiner (1976). N02-, N03- and low molecular weight organic acids in growth medium were
analysed by ion exchange chromatography (IonPac ATC1 anion trap column, IonPac AGll
guard column, analytical IonPac ASll 4 mm column, ASRS-II 4 mm suppressor auto
regenerating mode, and CD20 conductivity detector, Dionex, Olten, Switzerland). Elution was
performed with a gradient of NaOH 0.5 mM to NaOH 27.5 mM in 10 minutes at a flow rate
of 1 ml·min- I. N02- and N03- were estimated additionally with analytical test strips (Merck,
Darmstadt, Germany, concentration range: N02-: 0.1-3 g rI, N03-: 10-500 mg r l). TMA was
determined as described by Shen (1988). Total nitrogen (Ntot) was photometrically determined
with a Nitrogen Cell Test (Merck). The C and N content of the cells were determined with a
28
Pseudomonas putida TM 1
CHNS-O analyser (Carlo Erba Instruments, Milano, Italy). Cells used for C and N analysis
were washed three times with distilled water and lyophilised prior to analysis. CO2 and O2
were analysed by gas chromatography (gas chromatograph type GC-8A, Shimadzu Co.,
Tokyo, Japan) equipped with two parallel packed columns packed with molecular sieve 5 A
80/100 and Porapack Q 80/100, respectively (both from Brechbtihler AG, Schlieren,
Switzerland), the carrier gas was helium, detection was achieved by thermal conductivity.
TM, DM, MM and choline were measured by ion pair chromatography as described by Weiss
(1991) for the determination of choline except that the eluent composition was changed to a
ratio of 98 % hexansulfonic acid (2 mM) 1 2 % acetonitril. Samples were filtered and diluted
ten times (concentration < 40 mg r l) prior to analysis. The detection limit for the QAAs in
distilled water was 4 mg r l. However, in growth medium a ten-fold dilution of all samples
was required to reduce matrix component effects and consequently the detection limit in
growth medium was only approximately 40 mg r I.
RESULTS
Enrichment, isolation and identification. Enrichment for TM-degrading microorganisms
was performed in batch and continuous culture at 30 QC and a pH of 7 using SM
supplemented with TM as the sole source of carbon, energy and nitrogen at a concentration
range of 50 to 250 mg r I. Continuous enrichment cultures were run in small chemostats
(100 ml volume) at D - 0.04 h- I and a feed concentration of TM of 150 mg r I. Both batch and
continuous cultures were inoculated with soil (Duebendorf, Switzerland), river water (river
Chriesbach, Duebendorf, Switzerland) or activated sludge from two wastewater treatment
plants (WWTP Duebendorf, Switzerland, and model wastewater treatment plant, SEAC,
Unilever Research, Port Sunlight, UK). Enrichment cultures showing positive growth
(turbidity) were diluted appropriately, plated on SM/TM agar plates and colonies of different
morphology were picked. In this way, two different strains growing on TM, originating from
continuous enrichment cultures, were isolated in pure culture. Strain TM 1 was selected for
further investigations due to its high maximum specific growth rate of 0.40 h- I during growth
on TM and its good handling properties for laboratory work. Strain TM 2 was growing much
slower on TM (/Lmax(TM) - 0.16 h- I) and only about 40 % of initially provided carbon (as TM)
29
Chapter 2
was used in batch cultivation, whereas TM 1 utilised 80-90 % of the initially supplied TM
carbon.
By sequencing the 16S-rDNA gene isolate TM 1 was identified as a strain of Pseudomonas
putida (in the following text referred to as P. putida TM 1). An alignment of the
P. putida TM 1 sequence to sequences in the databases of EMBL (European Molecular
Biology Laboratory) and RDP (Ribosomal Database Project, MAIDAK et. al. 1999) resulted
in a similarity of 99.9 % (best hit) with one base pair difference to strain P. putida DSM 291T.
The difference was located at position 1136 (E. coli nomenclature) of the 16 S-rDNA
sequence, with adenine in P. putida TM 1 and thymine in P. putida DSM 291 T. Affiliation of
P. putida TM 1 and DSM 291T to true fluorescent Pseudomonads was verified by plating
them on Gould SI iron-deficient medium, on which the production of green-fluorescent
pigments was detected for both strains.
Morphology of colonies and cells. The cells of P. putida TM 1 were motile and elliptical in
shape with variable size (0.5-0.9 x 1.2-2.4 /-Lm) and two or three polar flagella (Figure 2.2a,d).
All cells were morphologically identical within the size range mentioned. Aggregates of a few
up to about a hundred cells were formed while growing with TM as sole source of carbon and
nitrogen, typically in the initial batch growth phase at low cell densities. Separation of
clustered cells was observed later at higher cell densities. The morphology of strain P. putida
TM 1 is clearly different from that of cells of the reference strain P. putida DSM 291T, which
appeared as slim rods with parallel cell walls (Figure 2.2b). The KOH and the L-alanine
aminopeptidase method indicated a Gram-negative nature of strain TM 1, whereas the Gram
staining gave equivocal results. Confirmation of the Gram-negative cell wall was obtained by
electron microscopy (Figure 2.2c).
On agar plates containing MSITM, P. putida TM 1 formed irregular colonies with a clearly
defined border of a beige, shiny appearance (diameter < 0.5 mm) after one day of incubation
turning into brown, mat colonies (diameter - 1 mm) after the second day. Longer incubation
times lead to larger, volcano-like colonies of up to 2 mm diameter without changes in color.
On TSA plates the colonies were circular with clearly defined borders (diameter 1 mm) and
appeared dark brown and mat after one day of incubation. Longer incubation times did not
alter the shape or the color and only the size increased to about 4 mm after three days.
30
Pseudomonas putida TM 1
d
-.,'-- fill;
1,/1..,\,
\,
,,-
'" ,.,,~ ."....
Figure 2.2. Morphology of isolate P. putida TM 1 (a) and of the reference strain P. putida DSM 291 T
(b) seen in the light microscope. Both strains were pregrown on tryptic soy broth. Electron microscope
pictures of thin sectioned (c) and negatively stained (d) cells of P. putida TM 1.
Colonies of P. putida DSM 291 T growing on TSA plates were irregular with a distinct border
after one day of incubation and of light brown, mat color with a diameter of about 1.5 mm.
After 2 to 3 days colonies became larger (up to 5 mm of diameter) and fringed, without
changing color.
Nutritional and biochemical properties. For the determination of the nutritional and
biochemical properties of P. putida TM 1 a variety of selected organic compounds were tested
and the commercially available test systems API 20 NE, API 50 CH and Biolog GN plates
were used. API 20 NE and API 50 CH and some selected tests were executed as well with
P. putida DSM 291 T.
31
Chapter 2
In growth tests using liquid SM, P. putida TM 1 was found to grow exclusively with the QAA
TM, which was used for its isolation, whereas DM and MM did not support growth. Choline,
the natural structurally related compound to TM, and all metabolites of its degradation
pathway, namely betaine, dimethylglycine, sarcosine, and glycine (Kortstee, 1970) were
utilised for growth by P. putida TM 1. Out of the additionally tested CIN-containing
compounds, only ethanolamine supported growth whereas methylethanolamine,
methyldiethanolamine, dimethyl-2-propanolamine, triethanolamine, ethylendiamintetraacetic
acid, and nitrilotriacetic acid did not. Cz-compounds such as ethanol, acetate or glyoxylate as
well as the fatty acids propionate and octanoate served as growth substrates, too, whereas the
Cl-compounds formate, methanol, trimethylamine, dimethylamine and monomethylamine did
not support growth. Reference strain P. putida DSM 291 T failed to grow with the QAAs TM,
DM and MM even if an additional nitrogen source (NH4Cl) was added to the batch culture
medium. Except for TM, P. putida DSM 291 T showed the identical substrate utilisation
pattern as P. putida TM 1 did.
To test whether DM or MM can be cometabolised, the two quaternary amines (DM 350 mg rl
and MM 440 mg r\ respectively) were pulsed into a continuous culture of P. putida TM 1
growing on TM as sole source of carbon and nitrogen. However, neither DM nor MM was
used, and the two pulsed compounds were washed out following the theoretical washout
curve (Figure 2.3a,b). No toxic or inhibitory effects of DM and MM on growth with TM were
observed, indicating that the two quaternary ammonium alcohols do not interfere with the
uptake of TM.
The results of API 20 NE, designed for non-enteric bacteria, are shown in Table 2.1 and give
an impression of the basic metabolic abilities of P. putida TM 1. Although, according to the
manufacturer, this test should allow identification of isolates belonging to the genospecies
P. putida, strain TM 1 could not be identified as a member of P. putida by API 20 NE. The
reason for the failure to identify strain TM 1 was found in the test for the assimilation of
adipate. Whereas the isolated strain P. putida TM 1 was able to grow with adipate as carbon
source P. putida DSM 291 T was not.
32
Pseudomonas putida TM 1
3.0 300C
~
Cl
2.0 200 .sCD z..,.'" I
0 +0
..,.I
1.0 100 Z
<500
0.0 0
3.0 500d
400 CCl
2.0 .sCD 300..,.
Z'"0 +I
0..,.
200 I1.0 Z
100 :tI-
0.0 0
-1 0 1 2 3 4 5 6 7 8 9 10Time [h)
Cl
200 .sz
I
+ ..,.I
100 Z<5oo
Cl
200 .sz
I
+..,.I
100 Z<5oo
300
o300
<><> b<:to~~~
'()''C
.......... "i-.4t ••<> '0' .<>.~
<>
a<>.~<><>.
~c>v •
..•..... i·Q~<>.<>
~""
<>
M 0-2 0 2 4 6 8 10 12
Time [h)
0.0
3.0
2.0
3.0
2.0
CD..,.'"o
o1.0
CD..,.'"o
o1.0
Figure 2.3. Co-metabolic ability of P. putida TM 1 in the chemostat. A pulse of a) DM (350 mg r l),
b) MM (440 mg r l), c) ethanol (410 mg r l
), and d) TM (400 mg r l) was added to a continuous culture
of P. putida TM 1 growing with TM (2 g r l) as the only carbon, nitrogen and energy source,
D =0.1 h- I. Pulses were executed at time zero. OD546 (.), TM (0), DOC (<», N~+-N (L.),
theoretical washout curve (---).
Regarding the test results obtained with API 50 CH, which allows testing growth with 49
different carbohydrates, P. putida TM 1 exhibited very poor growth on all sugars. Only slight
growth was detected on 7 different sugars after up to 96 hours of incubation. Sugars
supporting growth of P. putida TM 1 and differences to P. putida DSM 291T are listed in
Table 2.1.
Biolog GN plates provide a substrate utilisation pattern of Gram-negative microorganisms by
testing activity (due to a pH-change) on 95 different carbon sources, including a variety of
carbohydrates, acids and amino acids. P. putida TM 1 showed metabolic activity on most
tested acids, whereas activity was poor on the tested carbohydrates (data not shown),
confirming the results obtained with API 50 CH and API 20 NE test kits.
33
Chapter 2
Table 2.1. Physiological and morphological properties of P. putida TM 1 as collected with API 20
NE, API 50 CH and selected additional tests. If carbon source was not nitrogen-containing, N14CI
was added to the growth medium (0.8 g.-I). Note: Growth with arabinose was positive for both strains
when API 20 NE was used whereas P. putida DSM 291T exhibited no growth in test kit API 50 CH.
This difference may be caused by the different media used and the different concentrations of
provided substrate in these two tests and it points to the limitation of their interpretation. +...positive
reaction, definite growth; ± ...slight growth; -...no reaction, no growth; ( )...result assessed after 48
hours of incubation.
+ +-(+) -(+)-(±) -(±)-(±) -(±)-(±) -(±)-(±) -(±)
+ ++ +++ ++ ++ +
Properties tested
Shape:Motility:
API20NE:Reduction of nitrate to nitriteReduction of nitrates to nitrogenIndole productionAcidificationArginine dihydrolaseUreaseHydrolysis of B-glucosidaseProtein hydrolysisPresence of B-galactosidaseAssimilation of:
GlucoseArabinoseMannoseMannitolN-Acetyl-glucosamineMaltoseGluconateCaprateAdipateMalateCitratePhenyl-acetate
Presence of cytochrome oxidase
P. putida TM 1
ellipsoidmotile
+
P. putida DSM 291T
rodmotile
+
API 50 CH:GlycerolL-ArabinoseRiboseD-GlucoseD-FructoseGluconate2-Keto-gluconateAll 42 other sugars
±±±± +± ±± +± ±
Assimilation of:TMDMMMCholine
+
+
34
+
Pseudomonas putida TM 1
A comparison of protein patterns with SDS-PAGE was performed for both strains, P. putida
TM 1 and P. putida DSM 291T. P. putida TM 1 cells grown with TM or choline as sole
source of carbon and nitrogen exhibited different protein patterns suggesting that different
enzymes are involved in degradation of TM and choline despite their structural similarity
(Figure 2.4). Striking in the SDS-PAGE from both, TM- and choline-grown cells, was the
distinct protein band at about 68 kDa which was less pronounced in TSB-grown cells
(Figure 2.4). SDS-PAGE with cells of P. putida TM 1 and P. putida DSM 291T grown on
identical medium with same carbon source, namely either choline or TSA, resulted in
identical protein patterns for both substrates used (data not shown).
grown on grown on grown onTM choline TSB
kDaStandardHAA401
TM 1 TM 1 TMl
200.0
97.4
68.0 -
43.0 -
18.4 -
14.3 -
Figure 2.4. SDS-PAGE of total cellular proteins obtained from P. putida TM 1 grown with TM,
choline, and TSB. Arrows point out major differences in cellular protein between TM-grown and
choline-grown cells.
Growth characteristics. Optimum growth conditions of P. putida TM 1, growing with TM
as sole source of carbon and nitrogen, were found to be 30 QC and pH 6.5 to 8 resulting in a
f.1max of 0.40 ± 0.02 h-1. Growth was still observed at 4 QC, but not at 37 QC. At pH 9 the
specific growth rate was reduced to about 75 % of f.1max measured at optimum conditions,
whereas at pH 5 even a reduction to 20 % was observed. The C- and N-content of
35
Chapter 2
P. putida TM 1 biomass grown with TM was 45 % C and 12 % N of cell dry weight, which is
within the limits normally observed (Egli, 2000).
The ability of isolate TM 1 to degrade TM was found to be stable since cells grown on TSB
for more than one week (transferred to new medium every day, totally at least 60 generations)
were able to grow immediately with TM after transferring them back to TM-containing
medium again.
Typical batch growth of P. putida TM 1 with TM in synthetic medium is shown in Figure 2.5.
The culture grew exponentially in batch cultures. However, after reaching about 60 to 70 % of
final ODs46 an immediate decrease to a new constant growth rate of about 50 to 60 % of the
initial growth rate was observed. Excess nitrogen from TM was released in the form of
ammonia and the release was found to be proportional to the biomass increase. DOC and TM
decreased proportionally to the increase of biomass indicating that the growth yield during the
entire batch growth stayed constant.
0.8 600 6
500 5
0.6 -------------~-
400 ~=- 4 ~
Cl
oS Clco oS'<ton 0.4 300 ()0 3 z0 0 I
0 +'<t
~:r:
200 I- 2 z
0.2 -~.~.-. __._--
100
0.0 0 00 2 4 6 8 10 12 14
Time fhl
Figure 2.5. Batch growth of P. putida TM 1 with TM as sole source of carbon, energy, and nitrogen.
T =30 QC, pH =7.0 ± 0.1; OD546 (.), TM (0), DOe (0), NH/-N (6.).
36
Pseudomonas putida TM 1
The carbon and nitrogen balances in batch and continuous culture are given in Table 2.2. In
batch culture the residual TM concentration after exponential growth was below the detection
limit, i.e. below 40 mg r l. The final DOC concentration typically reached about 25 mg r 1
carbon (not including carbon originating from EDTA in the SM). Growth with TM in batch
culture was characterised by a low growth yield YX/c of 0.73 mg dry weight per mg carbon,
indicating a high level of carbon combusted to CO2 (60 %). Unusually, pH decreased
significantly (ca. 0.5 units) during growth despite the buffering and although excretion of
ammonia occurred. Since no organic acids such as acetate, formate, oxalate, or other small
strong acids could be detected the pH decrease was probably due to the high amount of CO2
produced during growth.
Table 2.2. Carbon and nitrogen balances for P. putida TM 1 growing with TM as sole source of
carbon, energy and nitrogen in continuous and batch culture. Mean values ± standard deviation are
given. Values reported are based on at least 4 samples in continuous culture and at least 2 samples in
batch culture.
Carbon balance Continuous culture Batch culture
mgC'h,1 % mgc·r l %
Input: 118.2 ± 3.5 100 279.5 100
Output (total): 117.2 ± 5.8 99 279.5 ±l.l 100
Biomass 27.9 ± 0.9 24 85.3 ± 0.5 31
DOC 6.4 ± 0.2 5 25.7 ± 0.6 9
CO2 dissolved 2.1 2
CO2 80.8 ± 5.7 68 168.5 ± 0.8 60
Input-Output: 1.0 ± 6.8 1 0.01) 01)
Nitrogen balance Continuous culture Batch culture
mgN'h,1 % mgN·r l %
Input: 23.1 ± 0.7 100 54.1 100
Output (total): 22.1 ± 0.7 96 54.1 ± 0.6 100
Biomass 7.2 ± 0.2 31 21.5 ± 0.1 40NH4+ 14.7 ± 0.4 64 27.9 ± 0.4 52
DON 0.2 ± 0.5 1 4.7 ± 0.4 8
Input-Output: 1.0 ± 1.0 4 0.01) 01)
1) Total CO2 release and DON were calculated as difference of measured input minus measured outputparameters, hence input - output = 0
In carbon-limited continuous culture (2 g r 1 of TM corresponding to about 1.075 gC r\ the
steady-state residual TM concentration at D = 0.1 h-1 was below the detection limit of
37
Chapter 2
40 mg r 1. Residual DOC concentration in the culture amounted to 56 mg r 1 which is a rather
high level of residual carbon in a C-limited continuous culture indicating the excretion of
carbonaceous products. However, no small organic acids (acetate, formate, oxalate or others)
were detected in the culture liquid. The growth yield in chemostat culture YXlC was
determined as 0.54 mg dry weight per mg carbon. This is significantly lower (26 %) than the
yield measured in batch culture. The decreased growth yield in the chemostat might be due to
a high requirement of energy for maintenance at low dilution rates but this still remains to be
elucidated. In any case, the excretion of products (DOC) was too low to explain the measured
difference in growth yields.
The majority of surplus nitrogen from TM was released as N~+ both in batch (- 85 %) and in
continuous culture (- 98 %) and neither N02-, N03- nor trimethylamine was excreted by
P. putida TM 1. Residual DON in continuous culture was 1 % of totally provided nitrogen
(from TM) and in the range of the standard deviation. In batch culture, residual DON
amounted to about 8 % of the totally provided nitrogen, the majority probably being
contained in TM. Since detection limit for TM was 40 mg rI, it was not possible to
characterise/identify the remaining DON.
Because in nature and wastewater treatment plants mixed substrate conditions are the rule
rather than the exception growth of TM 1 was investigated in batch and C-limited continuous
cultures using mixtures of substrates, too. In batch culture, no diauxic growth of
P. putida TM 1 was observed when the bacterium was cultivated in media containing TM as
sole source of nitrogen using either acetate or ethanol as an additional carbon source. The
different substrates were utilised simultaneously and the release ofN~+ into the medium was
highly reduced due to the additional formation of biomass (data not shown). An increase of
the growth rate was observed for the combination of TMlacetate, whereas no change in
growth rate was detected for TMlethanol. When pulsing ethanol (400 mg r 1) into a
continuous culture of P. putida TM 1, growing with TM (input 2 g r 1) as sole source of
nitrogen, ethanol was used immediately resulting in an increase of biomass concentration and
a decrease of the concentration of~+ in the culture medium. As a reference experiment,
TM (400 mg r 1) was pulsed to the same continuous culture resulting in an immediate
formation of additional biomass and an increase of N~+ concentration in the medium
proportional to the increase of biomass (Figure 2.3c,d).
38
Pseudomonas putida TM 1
DISCUSSION
Quaternary ammonium surfactants (cationic surfactants) have now been used in fabric care
for more than thirty years with an estimated world production of some 350'000 tonnes per
year (Van Ginkel, 1995). Therefore, several studies concerning the environmental properties
of these compounds have been performed (summarised in: Callely et aI., 1977; Krueger et aI.,
1998; Van Ginkel, 1995). Most investigations have focused on the fate, toxicity and
biodegradability in natural or simulated natural systems such as OECD test procedures. A
limited number of studies on the isolation of bacteria able to degrade quaternary ammonium
surfactants have been reported so far (Dean-Raymond & Alexander, 1977; Nishihara et aI.,
2000; Van Ginkel et aI., 1992). Microorganisms were isolated from activated sewage sludge
or soil using the quaternary ammonium surfactants decyl-trimethyl-ammonium, hexadecyl
trimethyl-ammonium and didecyl-dimethyl-ammonium as sole source of carbon. However,
the degradation of these compounds was either incomplete (excretion of tri- or dimethylamine
into the growth medium) or restricted to a consortium of at least two different
microorganisms. None of the isolated bacteria were able to completely degrade the alkyl part
and the methylated nitrogen part. Based on these studies and other investigations on the
biodegradation of a series of anionic, non-ionic, amphoteric and cationic surfactants Van
Ginkel (1996) concluded that complete degradation of most surfactants has to be achieved by
consortia of microorganisms. He suggested that only a few surfactants, i. e. alkane
sulphonates, alkyl sulphates and alkylamines are completely degraded by single
microorganisms.
In contrast, the P. putida TM 1 strain described here was able to completely degrade both the
Cl groups as well as the N-linked alkyl part (i.e. not the fatty acid part) of TM. The nitrogen
contained in TM was used as nitrogen source and excess nitrogen was excreted in the
mineralised form of NH/. No trimethylamine was found in the culture medium. It appears
that the methyl groups in TM were not incorporated into biomass but rather oxidised to CO2.
This speculation is supported by the low yield and high percentage of produced CO2 during
growth with TM as well as by the failure of P. putida TM 1 to grow with Cl compounds.
In previous studies nitrogen was always added to the media as an~+ salt in addition to the
nitrogen supplemented with the quaternary ammonium compound. Hence, no selection
pressure was applied for microorganisms able to use the nitrogen provided with the
39
Chapter 2
quaternary ammonium surfactant. This was perhaps a reason for the failure of previous
studies to isolate bacterial strains able to degrade quaternary ammonium surfactants to
completion.
In the literature, several degradation mechanisms of alkyl-trimethyl-ammonium compounds
have been postulated (Callely et aI., 1977; Van Ginkel, 1995):
A) Oxidation at the free end of the alkyl chain followed by stepwise degradation via beta
oxidation resulting in betaine, a natural widespread compound utilised by a variety of
ffilcroorgamsms.
B) Progressive demethylation of the nitrogen centre, then splitting off the nitrogen from the
alkyl chain as ammonia. The alkyl chain finally undergoes again beta-oxidation (fatty acid
metabolism).
C) Cleavage of the N-Calkyl bond providing trimethylamine and the aldehyde of the alkyl
chain. Whereas the alkyl chain again is degraded via common beta-oxidation,
trimethylamine is degraded by methylotrophic organisms.
Mechanism C), initial cleavage of the N-Calkyl bond, was proposed to be a general strategy of
microorganisms to gain access to the alkyl chains of quaternary ammonium compounds (Van
Ginkel, 1996). However, no speculation on the mechanism(s) responsible for the degradation
of TM by P. putida TM 1 can be made based on the experiments performed so far.
Considering the structural similarity of TM to the naturally related compound choline, one
also could speculate that the degradation of TM might occur in a similar way to choline or
that it might be channelled into the choline pathway, since the ability to degrade choline is
widespread amongst microorganisms (Bater & Venables, 1977; Ikuta et al., 1977; Kiene,
1998; Kortstee, 1970; Ohta-Fukuyama et aI., 1980; Shieh, 1964). However, neither one of the
organisms that we have isolated nor the reference strain P. putida DSM 291T (also capable to
grow with choline) was able to degrade TM or one of the other QAAs. Therefore, degradation
of TM may not necessarily be associated with choline degradation.
The majority of the previously isolated bacteria found to attack quaternary ammomum
surfactants were assigned to the genus Pseudomonas (Dean-Raymond & Alexander, 1977;
Nishihara et al., 2000; Van Ginkel, 1996; Van Ginkel et aI., 1992). Our strain was a true
fluorescent pseudomonad according to the current taxonomic definition (De Vos & De Ley,
1983). Despite the striking similarity of the isolated strain P. putida TM 1 to P. putida
40
Pseudomonas putida TM 1
DSM 291T based on 16S-rDNA sequencing (99.9 % corresponding to one different base pair)
P. putida TM 1 differs in substrate range and even in shape from the reference strain. This
demonstrates that ribosomal, genetic analysis is able to provide phylogenetic relationship,
indeed, but does not necessarily supply information on the specific metabolic ability of
isolates.
Further research will be focussed on the enzymatic degradation pathway of TM. This
information will be important with respect to the environmental behavior of these compounds
which may help in designing more biodegradable alternative quaternary ammonium
compounds.
ACKNOWLEDGEMENTS
We thank Dr. Ernst Wehrli, Laboratory for Electron Microscopy, ETHZ, for preparation of
the electron micrographs of strain P. putida TM 1 and I. Holderegger for CHNS-analysis.
41
Seite Leer /Blank leaf
QAA degrading bacteria
3. Isolation and characterisation of bacteria able to grow with quaternary
ammonium alcohols as sole source of carbon, energy and nitrogen
ABSTRACT
The quaternary ammOnIum alcohols (QAAs) 2,3-dihydroxypropyl-trimethyl-ammonium
(TM), dimethyl-diethanol-ammonium (DM) and methyl-triethanol-ammonium (MM) are
hydrolysis products of their parent esterquat surfactants, which are widely used as softeners in
fabric care. We isolated several bacteria growing with QAAs as the sole source of carbon,
energy and nitrogen. The strains were compared with a previously isolated TM-degrading
bacterium, which was identified as a representative of the species Pseudomonas putida
(Kaech & Egli, 2001). Two bacteria were isolated with DM, referred to as strains DM 1 and
DM 2, respectively. Based on 16S-rDNA analysis, they provided 97 % (DM 1) and 98 %
(DM 2) identities to the closest related strain Zoogloea ramigera Itzigsohn 1868AL. Both
strains were long, slim, motile rods but only DM 1 showed the floc forming activity, which is
typical for representatives of the genus Zoogloea. Using MM we isolated a gram-negative,
non-motile rod referred to as strain MM 1. The 16S-rDNA sequence of the isolated bacterium
revealed 94 % identities (best hit) to Rhodobacter sphaeroides only. The strains MM 1 and
DM 1 were able to grow exclusively with the QAA used for their isolation. DM 2 was also
utilising TM as sole source of carbon, energy and nitrogen. However, all of the isolated
bacteria were able to grow with the natural and structurally related compound choline.
43
Chapter 3
INTRODUCTION
Today, esterquat surfactants are used in large quantities as softeners in washing detergents.
The annual production of esterquats probably exceeds 100'000 tons worldwide with basically
over 99 % of the material applied in fabric care (Krueger et al., 1998). The three mainly used
esterquat surfactants are shown in Figure 3.1. They hydrolise rapidly, abiotically and/or
biocatalysed, when reaching surface water or sewage treatment plants. The products are the
corresponding fatty acids and quaternary ammonium alcohols (QAAs) 2,3-dihydroxypropyl
trimethyl-ammonium (TM), dimethyl-diethanol-ammonium (DM) and methyl-triethanol
ammonium (MM) as displayed in Figure 3.1 (Giolando et al., 1995; Hellberg et al., 2000;
Krueger et al., 1998; Puchta et al., 1993; Simms et al., 1992; Waters et al., 1991). The parent
esterquats and the QAAs have been investigated extensively in standard biodegradation tests
Esterquat surfactants
o
oJlR
"" +/'---/ 0/N~OJlR
DM-esterquat
hydrolysis
\fatty acid
hydrolysis
\fatty acid
Quaternary ammoniumalcohols (QAAs)
oJl hydrolysis HO__ -"" /,---/OH
HO /'---/0 R~ ~ 'N+~N+ 0 '\ /~
/ ~OJlR fatty acid MM OHMM-esterquat
Choline, naturally related compound
R: Fatty acid carbon chain derived from tallow, mainly Cl6 and C18
Figure 3.1. Structures of the three main commercially used esterquat surfactants, their hydrolysis
products, and the structure of the naturally related compound choline.
44
QAA degrading bacteria
(GECD, 1981) mimicking degradation in complex systems. Based on these tests, both, the
parent esterquats and the QAAs, are expected to be readily and ultimately biodegradable in
the environment (Giolando et a!., 1995; Krueger et a!., 1998; Matthijs et a!', 1995; Puchta et
al., 1993; Simms et al., 1992; Waters et al., 1991; Waters et al., 2000). Their structural
resemblance to the naturally widely occurring compound choline (Figure 3.1) suggests that
QAAs are degraded in a similar way and at rates comparable to that of choline. However, the
three QAAs showed very different degradation patterns and different degradation rates in
GECD die-away tests (Hales, 1998) despite their structural similarity. From the simple and
choline-like structure of QAAs one would also expect that many different microorganisms
would be able to utilise the compounds as sole source of carbon, energy and nitrogen and that
no consortium is required for their degradation. However, no choline degraders, including
reference strains from the German Culture Collection DSMZ (Deutsche Sammlung fUr
Mikroorganismen und Zellkulturen), were able to metabolise these QAAs (own results).
Moreover, up to now, no microorganisms degrading the three QAAs have been isolated and
consequently the catabolic pathways are not elucidated yet. In view of the widespread
application of the esterquats in laundry detergents, we have set out to isolate and enrich
strains able to grow with TM, DM and MM.
Here we report the isolation and characterisation of bacterial strains able to grow with the
QAAs DM and MM as sole source of carbon, energy and nitrogen. With DM two strains were
isolated, referred to as strain DM 1 and DM 2. Only one strain, capable to degrade MM,
designated as strain MM 1, was isolated. The isolate able to grow with TM (P. putida TM 1)
has been described previously in detail (Kaech & Egli, 2001).
MATERIALS AND METHODS
Chemicals. The quaternary ammonium alcohols (QAAs) (±)-2,3-dihydroxypropyl-trimethyl
ammonium (TM), dimethyl-diethanol-ammonium (DM) and methyl-triethanol-ammonium
(MM) were provided by Unilever (SEAC Safety and Environmental Assessment Center,
Unilever Research, Port Sunlight, UK) as the iodine salts. All other chemicals were purchased
from Fluka unless indicated.
45
Chapter 3
Isolation, growth and maintenance of organisms. Isolation, growth and maintenance of
organisms was performed as described by Kaech & Egli (2001).
Bacterial strains and storage. Zoogloea ramigera Itzigsohn 1868AL (ATCC 19623, I-16-M)
was obtained from the German Culture Collection (DSMZ GmbH, Braunschweig, Germany).
For short-term storage, all strains were plated on lO-fold diluted TSA or agar plates
containing SM and a selective carbon source. For long-term preservation all strains were
suspended in 30 % glycerol and stored at -80°C.
16S-rDNA analysis. 16S-rDNA sequencing was executed as follows: 16S-rDNA fragments
of the isolated strains were amplified by PCR. The primers used for amplification were
originally reported by Weisburg et at. (1991). They were slightly modified resulting in the
sequencies 6F (5'-GGAGAGTTAGATCCTGGCTCAG-3') and 1510R (5'-GTGCTGCAGG
GTTACCTTGTTACGACT-3'). Both fragments were recovered and purified from the PCR
mixture by using Qiaquick spin columns (Qiagen, Basel, Switzerland), and ligated to pGEM
T-Easy (Promega, Wallisellen, Switzerland). Transformation of the ligation mixtures into
Escherichia coli DH5a resulted in several different recombinant strains, from which plasmid
DNAs were isolated according to Sambrook et at. (1989). Plasmid inserts of the proper size
were sequenced on both strands by using a Thermosequenase Kit (Amersham, Little Chalfont,
UK) with IRD-800 labelled primers (MWG Biotech, Ebersberg, Germany). Sequence
transcripts were separated and analysed on a LiCOR 4000L automated DNA sequencer
(LiCOR, Lincoln, NE, USA). All sequences were compared to the EMBL data bank entries
(European Molecular Biology Laboratory, Heidelberg, Germany) by using the BLAST2
routine (Gish, 1996-1999).
Nutritional and biochemical properties. Utilisation of different carbon sources and selected
metabolic activities were investigated in various commercially available ready test kits such
as API 20 NE and API 50 CH (Biomerieux, Geneva, Switzerland) following the instructions
of the manufacturer. For API 50 CH SM was used, amended with N~Cl (0.8 g r l) as
nitrogen source. The utilisation of selected carbon compounds and combined carbon/nitrogen
sources was tested in liquid culture (5 ml in glass test tubes) at pH 7 by following turbidity
46
QAA degrading bacteria
using SM supplemented with the compound of choice (50-250 mg r 1). Controls without a
carbon source were used. All tests were inoculated with cells taken from colonies pregrown
on tryptic soy agar. Cultures were incubated at 30 QC in the dark and all tests were done in
triplicates.
The Gram-reaction was tested with the classical staining procedure as described by Drews
(1968), the KOH method (Buck, 1982) and by testing for the presence of L-alanin
aminopeptidase (Bactident strips, Merck, Darmstadt, Germany).
SDS-PAGE of cellular proteins was performed as described by Laemmli (1970). Cells used
for SDS-PAGE were grown in batch cultures. Samples were collected in the exponential
phase of growth and cells were harvested by centrifugation in Eppendorf tubes in a
microcentrifuge (ALC, Milano, Italy). The supernatant was discarded and the pellet stored at
-20 QC not longer than one month prior to SDS-PAGE.
Growth characteristics. Experiments for the determination of optimum growth conditions of
the strains were performed in batch cultures in Erlenmeyer flasks containing 100 ml of SM
and the carbon source of choice (500 mg r1). Batch media were inoculated with cells pre
grown in the same medium taken from the exponential growth phase. Incubation was done at
different temperatures and pH. For batch growth at different pH values (5, 7, 9) the pH of the
buffer stock solution was adjusted with either NaOH or HCI before adding it to the medium.
With the exception of DM 1, growth was followed by measuring optical density at 546 nm
(ODS46) with a glass cuvette (1 cm) in a spectrophotometer (Uvikon Kontron, ZUrich,
Switzerland). Strain DM 1 formed huge flocs during growth, therefore, biomass was followed
by measuring excreted N~+, which turned out to be proportional to biomass formation. This
was confirmed by comparing the amount of cellular protein formed (measured with the Bio
Rad protein assay, Bio-Rad Laboratories GmbH, Munich, Germany) with the N~+ excreted
during batch cultivations. All experiments were performed at least in duplicate; numbers
reported with standard deviations always from triplicates.
Determination of the maximum specific growth rate (/lmax) was performed using optimum
conditions and under nutrient excess in batch culture following ODS46 and~+ excretion,
respectively. For this purpose, cells were transferred repeatedly into fresh medium before they
had reached the late exponential phase until they exhibited a stable /lmax.
47
Chapter 3
Carbon and nitrogen balances were performed in batch culture using bioreactors of 1 I
working volume (Bioengineering AG, Wald, Switzerland) or Erlenmeyer flasks of 2 I volume
under optimum conditions. The concentrations of DM and MM used was in the range of 200
to 300 mg r I. Balances for the strains DM 1, DM 2 and MM 1 from two independent cultures
were calculated over the entire growth phase, whereas mean values of parameters (formed
biomass, C- and N-content of the cells, DOC, N~+) were based on at least triplicates
obtained from independent batch experiments. Samples were taken from the culture,
centrifuged at 15000*g in a microcentrifuge (ALC, Milano, Italy) for 5 min and the
supematant was either directly used or stored at -20°C prior to analysis of the following
parameters: DOC, DM, MM, N~+, N02- and N03-. Cell dry weight was measured by
collecting cells from a known culture volume on preweighed Nuclepore polycarbonate filters
of 0.2 /lm pore size (Sterico AG, Dietikon, Switzerland). Filters were washed once with
distilled water and dried at 120°C. DOC was measured with a Tocor 2 Carbon Analyzer
(Maihak, Hamburg, Germany). Carbon originating from EDTA in the medium was subtracted
from DOC for the carbon balance. N~+-N was determined by the indophenol method as
described by Scheiner (1976). N02-, N03- and low molecular weight organic acids in growth
medium were analysed by ion exchange chromatography (IonPac ATC1 anion trap column,
IonPac AGl1 guard column, analytical IonPac ASl1 4 mm column, ASRS-II 4 mm
suppressor auto-regenerating mode, and CD20 conductivity detector, Dionex, Olten,
Switzerland). Elution was performed with a gradient of NaOH 0.5 mM to NaOH 27.5 mM in
10 minutes at a flow rate of 1 ml min- I. N02- and N03- were estimated additionally with
analytical test strips (concentration range: N02-: 0.1-3 g rI, N03-: 10-500 mg rI, Merck,
Darmstadt, Germany). The C and N content of the cells was determined with a CHNS-O
analyser (Carlo Erba Instruments, Milano, Italy). Cells used for C and N analysis were
washed three times with distilled water and lyophilised prior to analysis. DM and MM were
measured by ion pair chromatography as described earlier for the determination of choline
(Weiss, 1991), except that the eluent composition was changed to a ratio of 98 %
hexansulfonic acid (2 mM) I 2 % acetonitrile. Samples were filtered and diluted ten times
(concentration < 40 mg r I) prior to analysis. The detection limit for the QAAs in distilled
water was 4 mg r I. However, in growth medium a ten-fold dilution of all samples was
48
QAA degrading bacteria
required to reduce matrix component effects and consequently the detection limit in growth
medium was only approximately 40 mg r l.
RESULTS
Enrichment, isolation and identification. Enrichment for QAA degrading microorganisms
was performed in batch and continuous culture at 30 QC and a pH of 7. SM was used,
supplemented with either DM or MM as the sole source of carbon, energy and nitrogen at a
concentration range of 50 to 250 mg r l. Continuous enrichment cultures were run in small
chemostats (100 ml working volume) at D - 0.04 h- l and a feed concentration of the substrate
of 200 mg r l. Both batch and continuous cultures were inoculated with soil (Duebendorf,
Switzerland), river water (river Chriesbach, Duebendorf, Switzerland) or activated sludge
from two wastewater treatment plants (WWTP Duebendorf, Switzerland, and model
wastewater treatment plant, SEAC, Unilever Research, Port Sunlight, UK). Enrichment
cultures showing turbidity were diluted appropriately, plated on selective SMlQAA agar
plates and colonies of different morphology were picked. In this way, three different strains
originating from continuous enrichment cultures were isolated in pure culture, two of them
using DM (strains DM 1 and DM 2) and one strain using MM (referred to as strain MM 1) as
the sole source of carbon, energy and nitrogen.
The strains DM 1, DM 2 and MM 1 were identified by sequencing the 16S-rDNA gene. For
DM 1 (EMBL: AJ440749) and DM 2 (EMBL: AJ440750) the alignment to the sequences in
the EMBL database (European Molecular Biology Laboratory, Heidelberg, Germany) by
using the BLAST2 routine (Gish, 1996-1999) provided identities (best hit, total 1448 base
pairs) to Zoogloea ramigera Itzigsohn 1868AL (EMBL: X74915) of 97 % and 98 %,
respectively. Best hit for the alignment of the MM 1 sequence (EMBL: AJ440751) was found
to be Rhodobacter sphaeroides (EMBL: X53854) with identities of 94 % (total 1434 base
pairs) only.
Morphology of cells and colonies. The shape, size and flagellation of the cells of strain
DM 1 and DM 2 was very similar. They were long, slim rods with a size of 0.6-0.9 x 2.0-3.5
p,m (Figure 3.2a,b). Both microorganisms were motile with one to several polar flagella. The
49
Chapter 3
attachment of the flagella was not at the polar end of the rod-shaped cells, but slightly shifted
to the long side (Figure 3.2b). On average, DM 2 cells were slightly longer and slimmer than
the cells of DM 1. While growing with DM, the cells of both DM 1 and DM 2 formed
inclusion bodies, pointing to the accumulation of polyhydroxyalkanoates (PHA) (Figure
3.2a). Cells of DM 1 formed huge aggregates (Figure 3.2c) during growth with all tested
substrates (see nutritional and biochemical properties). The aggregates often stuck loosely to
the sides of silicon tubes used for sampling in batch cultures. However, DM 1 did not possess
a visible zoogloeal, gelatinous matrix in the light microscope. Even after staining with 1 % of
crystal violet as described by Friedman & Dugan (1968), no exocellular material was evident.
Figure 3.2. Electron micrographs of thin-sectioned cells (a) and negatively-stained cells (b) of strain
DM 1. Arrow points to a storage granule, probably PHA. Circles point out the attachment sites of the
flagella of strain DM 1. Cells of isolate DM 2 (not shown) looked the same as those taken of DM 1,
except that the cells were slightly thinner and longer. c) Aggregates of DM 1 cells formed during batch
growth, as seen in the light microscope.
50
QAA degrading bacteria
In contrast, cells of DM 2 did not form any aggregation or flocs independent on the growth
substrate and cell density. Obviously, no matrices or floc-forming exopolysaccharides were
produced by this strain.
On agar plates containing TSA (10 % tryptic soy content), the colonies of strain DM 1 were
circular with a clearly defined border, and still very small after two days of incubation at
30 QC (diameter 1 mm). Colonies appeared light brown and slightly shiny. After six days, the
colonies reached a diameter of 2 mm. The border changed to slightly irregular shape and the
colonies turned brown with a narrow, bright rim. The entire colonies were movable when
pushed with an inoculation loop. If DM 1 was plated on SM/DM plates, the colonies were
white and shiny and reached a diameter of 1 mm at the most. Movability of the entire colonies
was detected on these plates, too.
Colonies of DM 2, grown on TSA plates (10 % tryptic soy content), showed the same
appearance as those of DM 1 after two days of incubation. In contrast to DM 1, DM 2
colonies were dark brown in the center of the colony and became continuously lighter towards
the border after six days of incubation. The maximum diameter after this time was 2 mm.
DM 2 colonies, grown on SM/DM plates, showed the same appearance as DM 1. In contrast
to DM 1, it was not possible to move them as a whole on the agar surfaces.
The type strain Z ramigera Itzigsohn 1868AL was plated on TSA for colony description, too.
The size of the colonies was identical to DM 1 and DM 2. After two days, the colonies were
1 mm in diameter and reached 2 mm after an incubation time of six days. The colonies after 2
days were circular, volcano-like, mat and of light brown colour with a dark brown, clearly
defined border. After six days of incubation, colonies turned dark brown in the centre with a
continuous gradient to a light brown border and the whole colonies were sprinkled with small,
dark dots. Movability of the entire colonies on agar plates was also observed. Since the type
strain Z ramigera Itzigsohn 1868AL did not grow on SM/DM plates, no colony description
can be made for this case.
The cells of isolate MM 1 were non-motile and without flagella. They were ovoid to rod
shaped and the size was 0.6-1.1 x 1.3-2.0 /lm (Figure 3.3a). MM 1 did not form any
aggregates during growth on all tested substrates. High amounts of inclusions bodies were
observed while growing with acetate (Figure 3.3b), whereas only few inclusions were found
51
Chapter 3
b
10/lm
Figure 3.3. a) Electron micrographs of thin-sectioned cells of strain MM 1. Anows point to storage
granules, probably PHA. b) Storage granules in cells of strain MM 1 (anows) grown with acetate, as
seen in the light microscope.
in cells growing with MM. Fluorescent staining of the inclusions with nile red pointed to
accumulation of a lipophilic storage compound, probably PHA.
Strain MM 1 showed poor growth on TSA plates (10 % tryptic soy content). When streaked
out on agar plates, best growth was observed in the region of the initially highest
concentration of the cells where a continuous film of cells developed. Only a few single
colonies were found. Single colonies reached a diameter of about 1 mm after three days of
incubation. They appeared as light brown, mat colonies with a clearly defined border and a
rough looking surface. After nine days, the colonies reached a diameter of about 2-3 mm in
diameter. The border turned irregular and the colonies were sprinkled with dark dots.
Based on the 16S-rDNA analysis all isolates belong to the group of Gram-negative microbes.
This was found as well with the L-alanine aminopeptidase method. Whereas the Gram
staining confirmed this result for isolates DM 1 and DM 2, equivocal results were found for
MM 1. The KOH method did not show the formation of gluey material (i. e. indicating a
Gram-negative cell wall) for either of the strains. For all isolates, confirmation of a Gram
negative cell wall was obtained by electron microscopy.
Nutritional and biochemical properties. For the determination of the nutritional and
biochemical properties of the isolated strains a variety of selected organic compounds was
tested and the commercially available test systems API 20 NE and API 50 CH were used. API
20 NE and some selected tests were executed as well with the reference strain Z. ramigera
Itzigsohn 1868AL.
52
QAA degrading bacteria
In growth tests using liquid SM and TM, DM or MM as the sole source of carbon, energy and
nitrogen, DM 1 and MM 1 were found to grow exclusively with the QAA used for their
isolation (DM and MM, respectively). In contrast, strain DM 2 was able to grow with DM as
well as with TM. Choline, the natural structurally related compound to the QAAs, and the
typical metabolites of its catabolism, i. e. betaine, dimethylglycine and sarcosine (Kortstee,
1970), were utilised for growth by all isolated strains. Glycine was used as growth substrate
by DM 1 and DM 2, but not by MM 1. From the additionally tested CIN-containing
compounds, ethanolamine served as growth substrate for all isolated strains, but only MM 1
was able to grow with methylethanolamine, too. Methyldiethanolamine, dimethyl-2
propanolamine, triethanolamine, ethylendiamintetraacetic acid, and nitrilotriacetic acid did
not support growth for any of the isolated strains. With the C2-compound ethanol, growth was
detected for DM 1 and DM 2, but not for MM 1. On the other hand, the C2-compound
glyoxylate supported growth for DM 1 and MM 1, but not for DM 2. Acetate, and the fatty
acids propionate and octanoate served as growth substrate for all three strains. The C1
compounds formate, methanol and monomethylamine did not support growth for any of the
isolated bacteria.
The tested reference organism Z. ramigera Itzigsohn 1868AL did not grow with TM, DM or
MM. With the naturally related compound choline, however, growth was detected. Since the
most closely related strain to MM 1 (based on 16S-rDNA sequence) was Rhodobacter
sphaeroides, MM 1 was also tested for anoxic, phototrophic growth which is found for all
representatives of the genus Rhodobacter (Dworkin et al., 1999-2002). However, MM 1 was
not able to grow under such conditions. Neither growth nor production of pigments was
observed with MM 1 incubated anaerobically in the light with acetate as the source of carbon.
Table 3.1 shows the results of test API 20 NE designed for identification of non-enteric
bacteria. The test gives an impression of the basic metabolic abilities of the different
microbes. However, the pattern did not allow an identification of the isolated strains.
Nevertheless, the test showed differences among all the isolated and the reference strain and,
therefore, allowed distinguishing between them.
53
Chapter 3
Table 3.1. Physiological and morphological properties of strains DM 1, DM 2, MM 1 and Z. ramigera
Itzigsohn 1868AL as collected with API 20 NE and other selected tests. If the carbon source supplied
for growth in assimilation tests was not containing nitrogen, NI4CI was added to the growth medium
(0.8 g r'). +...positive reaction, definite growth; ± ... slight growth; -...no reaction, no growth; nd ...not
determined.
Properties tested Isolate DM 1 Isolate DM 2 Z. ramigera Strain MM 1Itzigsohn 1868AL
Shape: rod rod rod rodMotility: motile motile motile non-motile
API20NE:Reduction of nitrate to nitrite + +Reduction of nitrates to nitrogenIndole productionAcidificationArginine dihydrolaseUrease +Presence of B-glucosidase + + + +Protein hydrolysisPresence of B-galactosidase ± + + +Assimilation of:
Glucose + + +Arabinose + + +Mannose + +Mannitol + +N-Acetyl-glucosamine + +Maltose + + +GluconateCaprateAdipateMalate +CitratePhenyl-acetate
Presence of cytochrome oxidase + nd
Growth with:TM +DM + +MM +Choline + + + +
54
QAA degrading bacteria
API 50 CH, which allows testing for growth with 49 different carbohydrates, provided the
following results for the isolated microorganisms. MM 1 did not grow on any of the sugars.
Slight growth of DM 2 was detected with 7 sugars, whereas DM 1 showed definite growth on
26 different sugars.
SDS-PAGE of the isolated strains, including P. putida TM 1, grown with the QAAs, provided
clearly different protein patterns (data not shown). No SDS-PAGE was performed with the
reference strain Z. ramigera Itzigsohn 1868AL because this strain was not able to grow with
any of the QAAs.
Since MM 1 showed 94 % identities only to Rhodobacter sphaeroides at the 16S-rDNA level
and did not exhibit the key properties found in all strains of Rhodobacter, no further
comparison was made using Rhodobacter sphaeroides.
Growth characteristics. Optimum growth conditions for isolate DM 1 growing with DM as
sole source of carbon, energy and nitrogen turned out to be 30 QC and a pH of 7, and resulted
in a fJ-max of 0.073 ± 0.007 h-1• Growth was still detected at 4 QC, but not at 37 QC. At a pH of
9, the growth rate was reduced to about 60 % of fJ-max and at pH 5 cells did not grow anymore.
The content of carbon and nitrogen in the biomass was 50 ± 3 % and 13 ± 1 % of the cell dry
weight, respectively.
Floc formation of strain DM 1 did not allow reliable determination of growth by OD
measurements. However, in batch cultivation, strain DM 1 grew exponentially (Figure 3.4a,b)
measured by excretion of NH/. DOC decreased to about 57 ± 2 % of the carbon
concentration provided initially as DM. 24 ± 2 % of the carbon were incorporated into
biomass and the remainder (19 ± 3 %, calculated as difference to 100 %) was most probably
released as CO2 (Figure 3.5). Proportionally to the biomass formation surplus nitrogen was
excreted as~+ and it amounted finally to about 21 ± 1 % of initially provided nitrogen in
DM. No nitrate or nitrite was detected in the culture liquid, some 29 ± 3 % was found
incorporated into the biomass, and the remaining nitrogen (50 ± 3 %) was present in the
culture supernatant as dissolved organic nitrogen (Figure 3.5). Strain DM 1 metabolised not
all of the initially provided DM. At the end of batch cultures, about 35 % of the initially
provided DM was left untouched in the culture liquid (Figure 3.4a). The reason for this
behavior was not caused by a shift of the pH, since the pH was controlled and adjusted if
55
Chapter 3
necessary during batch growth. Moreover, different initial concentrations in the range of 250
to 500 mg r l of DM in batch cultures provided the same outcome, and in the same synthetic
medium strain DM 2 was able to produce about twice as much biomass than strain DM 1.
Therefore, the incomplete utilisation of DM in batch cultures of strain DM 1 was most likely
not caused by a limitation of nutrients or trace elements in the SM. Furthermore, using higher
concentrations of other CIN sources, strain DM 1 as well as other strains were able to produce
much more biomass in the SM used without any indication of a nutrient or trace element
limitation.
Unutilised DM contributed to about half of the final DOC concentration (Figure 3.4a) and
made up all of final DON. This indicates that only carbon-containing metabolites and no
organic, nitrogen-containing compounds were excreted during growth. However, these carbon
compounds have not been identified so far. The growth yield in batch cultures was
approximately 1.1 ± 0.1 g dry weight per mg of carbon.
For strain DM 2, optimum growth conditions with DM as the sole source of carbon, energy
and nitrogen were found to be 25 QC and a pH of 7 - 9. The maximum specific growth rate
J-tmax was 0.078 ± 0.005 h- l. Whereas cells of DM 2 exhibited still growth at 4 QC, no growth
was detected at 37 QC and below a pH of 5. The carbon and nitrogen content of the cells was
46 ± 3 % and 12 ± 1 % of the cell dry weight, respectively. DM 2 grew exponentially in batch
cultures (Figure 3Ac). At the end of the exponential growth phase only 4 ± 1 % of the carbon
initially provided as DM were left. 41 ± 2 % of the carbon were assimilated into biomass and
the remainder (55 ± 3 %, calculated as the difference to 100 %) was combusted to CO2
(Figure 3.5a). Surplus nitrogen was excreted into the medium as .NI4+ and excretion was
proportional to the biomass increase. The final amount of NH/-N excreted was 45 ± 2 % of
the nitrogen initially provided as DM and 55 ± 2 % of the nitrogen was fixed into the biomass
(Figure 3Ac, 3.5b). No organic nitrogen (0 ± 4 %) was detected in the culture liquid at the end
of the exponential growth phase. Chemical analysis of DM confirmed that in batch culture the
DM provided was degraded to completion by strain DM 2 with a yield of 0.93 ± 0.05 mg dry
weight per mg of carbon. While growing with the QAA TM, DM 2 reached a J-tmax of 0.10 ±
0.01 h- l, which was much slower than the J-tmax found for P. putida TM 1 of 0040 ± 0,02 h- l
(Kaech & Egli, 2001).
56
QAA degrading bacteria
(a)
0.15 ..,..------------------r 300
<D~l!)
oo
0.10
0.05
+
+ +OM250
200 O'lE..........
150 ()oo
100 2o
50
3020Time [h]
10o0.00 +---~------.----.,.-----+0
40
(b)
0.15 --r----------------"'T"" 10
0.00 +--------,------.----....-----4
<D~l!)
oo
0.10
0.05
o 10 20Time [h]
30
8
6
4
2
o40
..--....,O'l.sz+'~
IZ
Figure 3.4. Batch growth of strains DM 1 (a, b), DM 2 (c, see following page) and MM 1 (d, see
following page) with DM (a, b, c) and MM (d) as sole source of carbon, energy and nitrogen.• DM,
o MM, 0 DOe, 0 OD546 mu, 6. NH/-N. Note: The detection limit of DM and MM with the
analytical method applied was about 4 mg r l. Since samples had to be diluted ten times before
analysis to reduce matrix effects, the conclusive detection limit raised to about 40 mg r l and probably
caused that the final DM level (zero) in Figure 3.4 c was reached before the beginning ofthe stationary
phase as detected by OD.
57
Chapter 3
(c)
0.35 300 20
0.30 250,........,
15,........,0.25
~
~
200 0>0>E
~ 0.20........ .s
0 150 g 10200.15 0 I
+'<t
100 ~I2
0.105
0.05 50
0.00 0 00 10 20 30 40
Time [h]
(d)0.5 300 5
0.4 0250 4
,........,~ ,........,
200 ~ ~
0>(0 0.3 E 3 0>'<t ........ EID
150 ()........
00 0 2
0 I
0.2 2 +'<t
100 ~ I~
2
0.1 50 1
0.0 0 0
0 2 4 6 8 10 12
Time [h]
58
QAA degrading bacteria
Optimum growth for strain MM 1 during growth in mineral medium with MM as a sole
source of carbon, energy and nitrogen was observed at 37 QC and at a pH of 7. The maximum
specific growth rate J.Lmax was 0.205 ± 0.006 h-1. Slow growth in batch culture (- 0.006 h- 1
)
was also detected at a temperature of 4 QC. At 38 QC, the specific growth rate was reduced to
about 65 % of J.Lmax and at 39 QC no growth was detected anymore. The carbon and nitrogen
content of the cells was 44 ± 3 % and 11 ± 1 % of the cell dry weight, respectively. In batch
culture strain MM 1 grew exponentially with a constant specific growth rate until entering the
stationary phase (Figure 3.4d). A relatively high proportion of 31 ± 4 % of the initially
provided carbon from MM was left in the medium in an unidentified form when the culture
entered the stationary phase. 37 ± 1 % was incorporated into the biomass and 32 ± 5 %
(calculated as the difference to 100 %) was combusted to CO2 (Figure 3.5a). The nitrogen
balance for growth with MM in batch culture (equivalent to 100 %) gave the following results
at the end of the exponential growth phase: 57 ± 1 % was fixed in biomass, 42 ± 2 % was
found as organic nitrogen in the culture liquid (DON, calculated as the difference to 100 %)
and only 1 ± 1 % was excreted as NH/-N (Figure 3.5b). No other nitrogen compounds like
N02- or N03- were found in the culture broth and MM disappeared completely during batch
growth (Figure 3.4d). It seems very likely that the high amount of residual nitrogen as DON
was a dead-end metabolite(s), which cannot be used anymore for growth. This was confirmed
with the data from samples taken four hours after the cells had entered the stationary phase,
where the same amounts of carbon and nitrogen were detected as at the beginning of the
stationary phase. The ratio of carbon to organic nitrogen in the remainder was 4.7 but it is
uncertain whether or not it is a single compound or several transformation products. Attempts
to identify the compounds by ion-pair and anion chromatography were not successful. The
yield during batch growth was 1.21 ± 0.07 mg of dry weight per mg of carbon.
59
Chapter 3
(a)
80
60%
40
20
oDM 1 DM 2 MM 1 TM 1
Strain
~ Biomass 11DOe D cO2
(b)
100 ...---,.---...---,---r-.----
80
60%
40
20
oDM 1 DM 2 MM 1 TM 1
Strain
~Biomass IIDON DNH/
Figure 3.5. (a) Carbon and (b) nitrogen balances of strains DM 1, DM 2 and MM 1 obtained from
batch cultures. Balances of P. putida TM 1 are shown for comparison. Total CO2 and DON were
calculated as difference of experimentally measured input minus experimentally determined output
parameters. Standard deviations are given in the text.
60
QAA degrading bacteria
DISCUSSION
Several studies on the environmental properties of quaternary ammOnIum compounds
(cationic surfactants) have been performed since they have been used for more than thirty
years in large quantities (Callely et al., 1977; Garcfa et al., 2000; Garcfa et al., 2001; Krueger
et al., 1998; Valles et al., 2000; Van Ginkel, 1995). However, most of these investigations
comprise data on the fate, toxicity and biodegradability in the environment and in simulated
natural systems. Only a limited number of studies reports on the isolation of bacteria able to
degrade quaternary ammonium compounds (Dean-Raymond & Alexander, 1977; Nishihara et
al., 2000; Van Ginkel et al., 1992). The compounds used for the isolation of microorganisms
from sewage treatment sludge were decyl-trimethyl-ammonium, hexadecyl-trimethyl
ammonium and didecyl-dimethyl-ammonium as sole source of carbon. These compounds
were either degraded incompletely with tri- or dimethylamines being excreted, or microbial
consortia were required for their complete mineralisation. None of the isolated bacteria was
able to degrade both, the alkyl based and the methylated nitrogen part of the molecule. Based
on a series of studies on anionic, non-ionic, amphoteric and cationic surfactants Van Ginkel
(1996) proposed that the complete degradation of most surfactants is performed by consortia
of microorganisms. He suggested, that only alkane sulfonates, alkyl sulphates and
alkylamines are entirely degraded by individual strains. The lack of the isolation of microbes
degrading quaternary ammonium surfactants to completion may be due to the fact that in all
those studies nitrogen was provided in the form of ammonium in addition to the nitrogen
containing surfactant. In this way, no selective pressure to use the nitrogen deriving from the
surfactant itself was imposed during enrichment and isolation.
In contrast, we report here the isolation of pure bacterial isolates that were able to grow with
the quaternary ammonium alcohols DM and MM as a sole source of carbon, energy and
nitrogen. Hence, all bacterial strains were able to degrade the carbon-containing chains (or at
least parts of it) as well as to use the nitrogen contained in the quaternary ammonium part of
the molecule for biomass formation. As an exception, strain DM 1 did not metabolise all of
the provided QAA DM in batch cultures. This behavior was not caused by a shift of the pH or
by a limitation of nutrients in the medium and therefore this phenomenon remained
unexplained. In contrast, DM 2 did completely degrade DM during batch cultivation with
only a small amount of residual DOC being left at the end of the growth and all surplus
61
Chapter 3
nitrogen released as ammonium. Strain MM 1 again metabolised all the provided QAA MM
in batch cultures but excreted high amounts of organic, nitrogen containing not yet identified
compounds that remained untouched in the stationary phase. The degradation of TM again
proceeded with a different pattern. Here, it was observed that P. putida TM 1 consumed all
the provided TM in batch culture and all surplus nitrogen was released as ammonium with a
considerable part of the initial carbon remaining as unidentified dissolved organic carbon
(Kaech & Egli, 2001). Obviously, each of the isolated strains degraded its QAA in a different
way despite the structural similarity of the QAAs. Even the two isolates DM 1 and DM 2,
which turned out to be closely related based on their 16S-rDNA sequence, showed a totally
different behavior during growth with the same QAA. Hence, one is tempted to speculate that
the degradation of the QAAs is achieved by different mechanisms. Surprisingly, DM 2 was
the only organism able to grow with more than one QAA. Considering the structural
similarity of the QAAs one would expect that microbial strains would be capable to degrade a
whole range of different quaternary ammonium alcohols. Obviously, this is not the case.
QAAs are structurally similar to choline. This compound is found ubiquitously in nature and
the ability to degrade choline is widespread amongst microorganisms (Kortstee, 1970;
Rosenstein et al., 1999; Shieh, 1964). Since all isolated bacterial strains were able to grow
with choline, one can speculate that the QAAs are metabolised in a similar manner as choline
is. However, neither of the bacteria we isolated with choline (results not shown) nor the
reference organisms Z. ramigera Itzigsohn 1868AL, nor P. putida DSM 291T
, both also able to
grow with choline, were able to degrade any of the QAAs. Therefore, the ability to degrade
choline does not go along with the ability to catabolise TM, DM or MM. Earlier, several
mechanisms have been proposed (Callely et aI., 1977; Van Ginkel et aI., 1992) for the
degradation of alkyl-trimethyl-ammonium compounds (not esterquats or quaternary
ammonium alcohols). In brief, the three alternative mechanisms that have been suggested
were 1) 00- and ~-oxidation of the alkyl chain resulting in betaine, which undergoes
progressive demethylation finally providing glycine; 2) Initial demethylation of the nitrogen
centre, then the splitting off of the nitrogen from the alkyl chain as ammonia and ~-oxidation
of the remaining alkyl chain; and 3) the splitting off of the alkyl chain from the quaternary
nitrogen atom with the release of the appropriate amine, the alkyl chain again undergoing ~
oxidation. However, from the results of this work no predictions can be made so far with
62
QAA degrading bacteria
respect to the mechanisms responsible for the degradation of the QAAs and a careful
investigation is needed for the elucidation of the metabolic pathway(s) involved.
In the literature, most isolated bacteria able to degrade quaternary ammonium compounds
were found to be representatives of the genus Pseudomonas (Dean-Raymond & Alexander,
1977; Nishihara et al., 2000; Van Ginkel, 1996; Van Ginkel et al., 1992) and recently we
reported the isolation of a Pseudomonas strain growing with TM (Kaech & Egli, 2001). We
report here the isolation of microorganisms that are closer related at the phylogenetic level
(16S-rDNA sequence) to the genus of Zoogloea and Rhodobacter than to Pseudomonas.
Hence, the degradation of quaternary ammonium compounds seems not to be an exclusive
trait ofmembers of the genus Pseudomonas.
The properties of the isolated strains DM 1 and DM 2 were compared to those found in the
literature for the closely related genus Zoogloea. Their morphology (size, shape, flagellation
and storage of granules) and the inability to grow at pH 5 corresponded entirely with the
properties found for all species of Zoogloea as reported by Dworkin et al. (1999-2002).
However, only strain DM 1 exhibited movability of entire colonies on agar plates, cell
aggregation, sticking of aggregated cells to tubes and the lack of a visible zoogloeal matrix as
was described for the closest related strain Z ramigera Itzigsohn 1868AL (Dworkin et al.,
1999-2002; Friedman & Dugan, 1968; Friedman et al., 1969; Unz, 1971). The absence of
these properties for strain DM 2, however, does not imply that DM 2 is not a Zoogloea
species, since many exceptions with respect to cell aggregation have been reported (Dworkin
et aI., 1999-2002).
Since strain MM 1 showed 94 % identical bases only to the closest related strain Rhodobacter
sphaeroides based on the 16S-rDNA and did not exhibit the key properties found for all
species of the genus Rhodobacter, MM 1 most likely does not belong to the genus of
Rhodobacter, but must be considered to be a member of a new genus.
We have shown here that a consortia of microorganisms is not required for the degradation of
quaternary ammonium alcohols deriving from esterquat surfactants and that the pattern of
degradation for individual quaternary ammonium alcohols differs quite distinctly. Further
research will be focused on the enzymatic degradation pathways of TM, DM and MM to
elucidate the enzymatic mechanisms and specificity of the degradation of QAAs. This
information will be important with respect to the environmental behavior of these compounds
63
Chapter 3
and may help in designing readily and ultimately biodegradable alternative quaternary
ammonium compounds.
ACKNOWLEDGEMENTS
Dr. Ernst Wehrli, Laboratory for Electron Microscopy, ETHZ is gratefully acknowledged for
preparation of the electron micrographs of the isolated strains. Thanks also go to I.
Holderegger for CHNS-analysis and Christoph Werlen for 16S-rDNA sequencing of the
isolated strains.
64
Microbial oxidation of MM
4. Microbial oxidation of methyl-triethanol-ammonium
ABSTRACT
Methyl-triethanol-ammonium (MM) originates from the hydrolysis of the parent esterquat
surfactant, which is used as softener in fabric care. The initial conversion of MM was
investigated in cell-free extracts of the bacterial strain MM 1 able to grow with MM as sole
source of carbon, energy and nitrogen. Enzymatic activity transforming MM was located in
the particulate fraction of strain MM 1. The oxygen dependent reaction occurred also in the
presence of phenazine methosulfate as alternative electron acceptor. As soon as one ethanol
group of MM was oxidised to the aldehyde, a cyclic hemiacetal (and its stereoisomers) was
built by intramolecular cyclisation. The third ethanol group of MM was oxidised to the
aldehyde and the carboxylic acid sequentially. However, no further oxidation was observed
for the cyclic hemiacetal. The structurally related quaternary ammonium compounds
dimethyl-diethanol-ammonium (DM) and choline were oxidised in the particulate fraction of
strain MM 1 as well. Since DM contains two ethanol groups, only the cyclic product (and its
stereoisomer) was formed. With choline, the expected primary and secondary oxidation
products betainealdehyde and betaine have been detected. The observed oxidation of MM,
DM and choline was also present in the particulate fraction of strain MM 1 grown with
acetate or choline. The oxidation of MM, DM and choline is most likely catalised by the
same, constitutively expressed membrane-associated oxidoreductase.
65
Chapter 4
INTRODUCTION
The quaternary ammonium alcohols methyl-triethanol-ammonium (MM), dimethyl-diethanol
ammonium (DM) and 2,3-dihydroxypropyl-trimethyl-ammonium (TM) are the three mainly
used head groups in esterquat surfactants, which are applied as softeners in fabric care
(Krueger et al., 1998). The parent esterquat surfactants hydrolyse rapidly, abiotically and/or
biocatalysed, when reaching surface water or sewage treatment plants to the fatty acids and
the quaternary ammonium alcohols (QAAs) (Hellberg et al., 2000; Krueger et al., 1998;
Puchta et al., 1993; Simms et al., 1992). The biodegradability of both, the parent esterquat
surfactant and the quaternary ammonium alcohol, has been investigated in standard OECD
biodegradation tests. Based on these tests they are considered as readily and ultimately
biodegradable (Krueger et al., 1998; Puchta et al., 1993; Simms et al., 1992; Waters et al.,
2000). Whereas the fatty acids are expected to biodegrade via the common fatty acid
catabolism (~-oxidation), the enzymes involved in the degradation of the quaternary
ammonium alcohol MM (and the other QAAs) are not yet known. Considering the
widespread application of these QAAs and the development and design of similar
compounds, it is important to know the strategies and the pathways of their biodegradation.
Therefore, we set out to enrich and isolate microorganisms able to grow with MM and to
investigate the enzymatic mechanisms that are responsible for its breakdown. We isolated a
microbial strain growing with MM as sole source of carbon, energy and nitrogen (see
Chapter 3). In this work, the initial step in the catabolism of MM was investigated in the cell
free extracts of the isolated strain. The protein fractions were also examined with respect to
the ability to transform the structurally similar compounds TM, DM, and choline.
MATERIALS AND METHODS
The quaternary ammonium alcohols (±)-2,3-dihydroxypropyl-trimethyl-ammonium (racemic
mixture), dimethyl-diethanol-ammonium and methyl-triethanol-ammonium were provided by
Unilever (SEAC Safety and Environmental Assessment Center, Unilever Research, Port
Sunlight, UK) as the iodine salts. All other compounds were purchased from Fluka, Buchs,
Switzerland, unless indicated otherwise.
66
Microbial oxidation of MM
Bacterial strains and cultivation. All experiments were performed with cells or cell-free
extracts of strain MM 1 isolated in our laboratory with MM as the sole source of carbon,
energy and nitrogen. An alignment of the 16S-rDNA sequence of strain MM 1 (EMBL:
AJ440751) to the sequences in the EMBL database (European Molecular Biology Laboratory,
Heidelberg, Germany) using the BLAST2 routine (Gish, 1996-1999) provided only 94 %
identities (best hit) of total 1434 base pairs to Rhodobacter sphaeroides (EMBL: X53854).
The physiological and morphological characterisation of strain MM 1 was reported previously
(Chapter 3).
Cultivation and storage of the strains was performed as described by Kaech & Egli (2001).
Preparation of cell-free extracts. Cells of strain MM 1 were grown in batch culture with 1 to
2 g r 1 of the indicated substrate. Cells were harvested in the late exponential growth phase by
centrifugation at 4 QC and 7000*g for 10 minutes (rotor: A 8.24, Kontron Instruments, Vietri
SuI Mare, Campania, Italy). Cells were washed once and resuspended after repeated
centrifugation with PB (50 mM). Before breakage of the cells - 20 mg r1 of DNAse I (EC
3.1.21.1, Aldrich, Milwaukee WI, USA) was added. Initially, 1,4-dithio-D,L-threitol (DTT,
2 mM) was added to the cell suspension as well, but later DTT was omitted, since it was
found to have no effect on enzyme activity in the protein fraction. The cells were broken by
two passages through a French press (Aminco, Urbana, USA) at 20000 psi. After each
passage the cell suspension and the French press were cooled with ice to 0 QC. The CE was
centrifuged for 30 min at 15000*g to remove unbroken cells and cell debris (rotor as above).
Subsequently, the cell-free extract was separated into a particulate (PF) and a supematant (SF)
fraction by ultra-centrifugation at 180000*g for one hour (rotor: TFT 65.13, Kontron
Instruments). The SF was removed and the pellet was suspended in PB (50 mM) and will be
referred to as the PF. Aliquots of the protein fractions were frozen at -20 QC until used in
assays. No reduction in activity was observed during freezing/thawing and storage. To
determine protein concentrations the Bio-Rad enzyme assay was used according to the
manual of the manufacturer (Bio-Rad Laboratories GmbH, Munich, Germany). BSA was
used as standard (Sigma-Aldrich Chemie GmbH, Steinheim, Germany).
67
Chapter 4
Enzyme activity assays. All assays were carried out in GC-Vials (Supelco Inc., Bellefonte
PA, USA) of 5 to 20 ml volume. Protein solutions were incubated at a pH of 7 and
thermostated at 30 QC under continuous stirring with a magnetic fly. The NMR samples were
prepared either by dilution with D20 (Aldrich, Milwaukee Wi, USA) or by carrying out the
assays with PB in D20. Protein concentrations used in the assays were always in the range of
0.4 and 1.0 mg mr I. To stop the enzymatic reaction, either aliquots removed from assays were
heated for 1 min in a water bath of 95 QC, or 1 M HCI was added (10 % of sample volume).
Before analysis, acidified samples were neutralised with 1 M NaOH (10 % of sample volume)
and the pH was controlled with test strips (Tritest pH 1 - 11, Macherey-Nagel, Dtiren,
Germany). IH NMR analysis indicated that using acid to stop the reaction did not change the
chemical composition of educts and products in samples. The oxygen dependence of the
reaction was monitored by carrying out experiments in sealed vessels with a continuous
nitrogen flow through the headspace. All solutions used in these experiments, except for the
protein stock solution, were bubbled with nitrogen for 10 min before adding them to the
assay. The electron acceptors NAD+, NADP+, PMS or INT were tested. NAD+, NADP+ and
PMS, final concentrations will be indicated at the appropriate place, were added to the assay
from stock solutions. INT was added to the reaction mixture from a saturated stock solution in
H20 (volume added to assays was 10 % of the total assay volume). For control experiments
protein solutions were heated in a water bath (- 95 QC) for 1 min prior to the addition of
substrate(s). Samples were stored at -20 QC prior to analysis. All enzyme specific activities are
given in nmol of converted substrate per min and per mg of protein (nmol min- I mg- I).
Chemical oxidation of MM. The oxidation of an ethanol group of MM to the corresponding
aldehyde was performed according to the method described by Dess & Martin (1983). BTF
was used as solvent (Ogawa & Curran, 1997) due to the poor solubility of MM in
dichloromethane. Saturated solutions of the oxidising reagent and MM in BTF were used for
the synthesis. After the reaction, residual MM and its oxidation products were extracted from
the BTF reaction mixture twice with H20. The combined water phase was freeze-dried and
the residual dissolved in D20 for NMR analysis.
68
Microbial oxidation of MM
Nuclear magnetic resonance spectroscopy. IH, 13C and 14N NMR spectroscopy were used
as analytical method to follow enzymatic activity and to elucidate the structure of the reaction
products. NMR spectra were recorded on a Bruker AMX-400 NMR spectrometer at 300 K
using a 5 mm broadband probe. The enzyme assays were directly performed in DzO buffer or
the samples spiked subsequently with DzO solution for the NMR lock signal. The NMR
experiments were performed without any further purification or filtration of the solutions. The
IH NMR (13C; 14N) spectra were recorded at 400.13 MHz (100.61; 28.9 MHz) with the
following parameters: 10.5 Ils (6.5; 13.5 Ils) 90° pulse lengths, appropriate number of
transients for reasonable SIN ratio, 8300 Hz (30300; 5800 Hz) spectral widths, 32k (64k; 16k)
data points, and 15 s (2.5, 1 s) relaxation delays. The I H NMR spectra were recorded with
presaturation of the water resonance using composite pulses (Bax, 1985). The 13C and 14N
NMR spectra were recorded using a proton decoupling field of 2.3 kHz during acquisition
(WALTZI6; Shaka et aI., 1983). The IH (13C) chemical shifts are given in parts per million
(ppm) relative to the signals of sodium 3-trimethylsilyl-tetradeutero-propionate (TSP) at
0.00 ppm (1.7 ppm).
The chemical shifts of the N-methyl groups of the substrates were used as internal reference
for the NMR spectra of the enzyme assays. The chemical shifts of 200 mM solutions of the
educts MM, DM, TM and choline in DzO were determined relative to TSP (18.7 mM) and are
given in Table 4.1.
The 14N chemical shifts of the same samples were determined relative to the signal of pure
nitromethane containing capillaries at 0.0 ppm and they were used as internal reference for the
NMR recorded during enzyme assays. The IH,14N HMQC (Bax et aI., 1983) experiments
were performed using the above mentioned 90° pulse lengths with the selection of a coupling
constant of 4 Hz showing the best results. The data was processed in the phase sensitive mode
to achieve better resolution in the 14N dimension for the methyl-nitrogen correlation signals.
IH,13C 2D correlation experiments were performed on a 5 mm broadband inverse probe with
z-gradient (100% gradient strength of 10 G cm-I) and 90° pulse lengths of 8.2 IlS eH) and
10.5 Ils (13C). The gradient selected HSQC (Davis et aI., 1992) (HMBC; Wilker et al., 1993)
experiments were performed with the selection of IH,13C coupling constants of 140 Hz
(5 Hz), gradient strengths of -40: 10 (15:9: 12), 2920 x 4800 Hz spectral widths with a carbon
decoupling field of 3.7 kHz for the HSQC experiments (GARP decoupling; Shaka et aI.,
69
Chapter 4
1985). The data matrices of 1024 x 256 were zero filled to 1024 x 1024. The HSQC-TOCSY
(Palmer et al., 1991) spectra were recorded with the selection of I JeH,13C) = 140 Hz and a
29 f.ls 90° pulse length for the TOCSY transfer with a total mixing time of 114 ms, applying
the above mentioned carbon decoupling conditions, data matrices and processing conditions.
The NOESY (Jeener et a!., 1979) spectra were recorded with data matrices of 1024 x 256
(spectral widths of 2400 x 2400 Hz) with presaturation of the water resonance during the
relaxation delay (2 s) and the mixing time (800 ms).
Quantitative determinations were performed using the IH NMR signals of the CH3-groups of
the corresponding compounds. The sum of all CH3-group integrals of a spectrum was set to
100 % and the amount of the educts and products were calculated according to their relative
intensities.
Table 4.1. I R, l3C and 14N chemical shifts of MM, DM, TM and choline (200 mM) in DzO in ppm and
coupling constants J( l3C,14N) in Rz.
Position MM DM Choline TMN-CH3 OeH) 3.26 3.23 3.21 3.24
O(13q 54.6 56.7 58.4 58.7IJ(13C,14N) 3.7 3.8 3.9 3.9
N-CH2 OeH) 3.68 3.60 3.52 3.43/3.48
oe3q 69.0 70.8 72.0 72.7IJ(13C,14N) 2.6 2.9 3.1 3.2
CH2-OH OeH) 4.07 4.07 4.07 3.60O(13q 59.7 59.9 60.1 70.0
CH-OH OeH) 4.30O(13q 68.2
Oe4N) -322.5 -328.1 -333.8 -334.1
Cation chromatography. Disappearance of MM in enzyme assays was also followed by
cation exchange chromatography using the following equipment: IonPac CTC-1 cation trap
column, IonPac C012 guard column, analytical IonPac CS12 4 mm column, CSRS-ULTRA
4 mm suppressor (external mode with tetrabutylammoniumhydroxyde 50 mM), and CD20
conductivity detector (Dionex, Olten, Switzerland). Elution was performed with HCI 15 mM
at a flow rate of 1 ml min- I. MM eluted 10 minutes (peak width - 2 min) after sample
injection.
70
Microbial oxidation of MM
Oxygen uptake measurements. Oxygen consumption was measured with a Clark-type
oxygen electrode (Rank Brothers, Cambridge, UK). Respiration was recorded before and after
the addition of MM to 1 ml of the protein solution in PB (0.4 - 1.0 mg protein mrl of PF) in a
closed reaction vessel at 30°C. To monitor the effect of KCN on Oz consumption, KCN was
added to a final concentration of 5 mM to a running assay with MM after a running time of 5
minutes. To detect possible HzOz production, catalase (EC 1.11.1.6, from bovine liver, Fluka,
Buchs, Switzerland) was added to an assay with MM after running for 5 minutes. The
presence of catalase in the PF was tested by adding HzOz to the protein mixture in the oxygen
electrode (without the substrate MM).
RESULTS
Enzymatic consumption of MM in cell-free extracts of strain MM 1. In enzyme assays,
disappearance of the quaternary ammonium alcohol MM was observed in cell-free extract
(CFE) and in the particulate fraction (PF) of MM 1 cells grown with MM as sole source of
carbon, energy and nitrogen (Figure 4.1). No consumption of MM was found in enzyme
assays performed with the soluble fraction (SF) and in control experiments with inactivated
extracts (CFE, SF and PF). Thus, the observed disappearance of MM must be mediated by
membrane-associated enzymes. Since the lH NMR spectra of reaction mixtures with the CFE
and the PF showed identical behaviour (disappearance of substrate and formation of the same
products), only the PF was used for further studies of the initial degradation step.
Simple stirring of the reaction mixture in open vials was sufficient to maintain MM
consuming enzyme activity and no additional reaction partners were required. The addition of
NAD+ and NADP+ to a final concentration of 2 mM did not enhance this activity. However,
under nitrogen atmosphere only a background activity for MM was detected (Figure 4.2,
initial slopes). This suggests an oxidative reaction to be responsible for the conversion of
MM. Oxygen consumption was confirmed and followed in a Rank oxygen electrode.
Immediate interruption of oxygen uptake was observed by addition of KCN to a running
assay. No HzOz was produced during the reaction as often found for oxidases (transferring the
electrons directly to oxygen), since the addition of catalase to an assay did not lead to oxygen
production. Moreover, no catalase activity was present in the PF, since addition of HzOz to the
71
Chapter 4
PF did not result in oxygen production either. These findings suggest that the electrons are
transported via the respiratory chain present in the particulate fraction. However, this is no
final proof, since cyanide might not only have interacted with terminal oxidases but also with
the MM-consuming enzyme.
As alternatives to oxygen the electron acceptors PMS, NAD+ and INT were tested under
anaerobic conditions. If the transport of the electrons to the terminal electron acceptor
occurred via the respiration chain one would also expect INT to be able to substitute for
oxygen under anaerobic conditions, since INT is known to act as an acceptor for electrons
deriving from the respiration chain. PMS, on the other hand, is known to act as a redox
mediator for many redox enzymes for which the natural redox mediating compounds are not
known. After incubation of enzyme assays for 3 h under N2 (di-nitrogen), the electron
acceptors were added to the reaction mixture and the change in MM concentration was
followed by IH NMR spectroscopy (Figure 4.2). PMS was the only reactant acting as
alternative electron acceptor. Although INT was not used as electron acceptor, this does not
exclude that the electrons are transported via the respiration chain.
1401201008060
........."'0... _
-0-- 0 0- --- ---0
4020
1.2
1.0
0.8~
.2l 0.6:2:2
0.4
0.2
0.00
Time [min]
Figure 4.1. Disappearance of MM (0) in an enzyme assay with particulate fraction of strain MM 1
grown on MM as sole source of carbon, energy and nitrogen, as determined by cation
chromatography. Activity of the particulate fraction based on the initial slope (solid line) was
260 nmol min- l mi l. The outlier indicated in brackets was not used to determine the initial slope.
72
Microbial oxidation of MM
Up to an initial MM concentration of 2 g r l, the substrate was completely consumed and
transformed into products within 2 h of incubation. Increasing the initial MM concentration to
6 g r l and up to 15 g r l, the amount of consumed MM decreased to 82 % and 12 %,
respectively. lH NMR data of an assay with 6 g r l (Figure 4.3) resulted in a specific rate
(initial slope) for MM disappearance of 324 nmol min- l mi l and from the cation
chromatography data of an assay with 1 g r l of MM an initial slope of 260 nmol min- l mg- l
can be estimated (Figure 4.1).
0.6 -r---------------------,
6543Time [h]
2
0.0 +---,----,----r------,r----.---lo
0.4
0.1
0.2
,g 0.3::2:::2:
Figure 4.2. Consumption of MM (0.5 g r l) in the particulate fraction of strain MM 1 (0.43 mg ml- I of
total protein) under anaerobic conditions as determined by IH NMR spectroscopy. After three hours,
the electron acceptors (0) PMS (4.5 mM), (I::::.) NAD+ (3.2 mM), or (D) INT (0.5 ml from a saturated
stock solution in water to 4.5 ml of the reaction mixture) were added to independent enzyme assays
(arrow). Only PMS stimulated activity significantly. A background level of specific enzyme activity of
approximately 34 nmol min- I mg- I was always detected in experiments under anaerobic conditions
(indicated by the initial slopes).
73
Chapter 4
Substrate specificity. The structurally related compounds 2,3-dihydroxypropyl-trimethyl
ammonium (TM), dimethyl-diethanol-ammonium (DM) and choline were tested in the
particulate fraction of MM 1 (cells grown with MM) under the same experimental conditions
as those used for MM. Whereas the PF exhibited activity with DM and choline as well, no
consumption of TM was observed. However, the specific activity with DM
(48 nmol min- I mg-I) and choline (23 nmol min- I mg-I) was much lower than with MM
(324 nmol min- I mg-I). Figure 4.3 shows enzyme assays with MM, DM and choline with
initial concentrations of 6 g r l and a protein concentration of 0.4 mg mr l. As for MM, the
activities with DM and choline were 02-dependent, only a slight residual activity was found
under di-nitrogen atmosphere, and PMS also acted as alternative electron acceptor. It should
be pointed out that although strain MM 1 was unable to grow with DM, this QAA was
transformed in the PF of MM-growing cells.
7
6
- 5.9(J)c
4(5.!:U
~ 30
~ 2~
----------------1)
[0]... ... ... ..., , , ,
"0.............0.
--- .... _------------'0
1084 6Time [h]
2
O+----,.------,,....-:l<.-----..----r-------!
o
Figure 4.3. Substrate specificity of the particulate fraction of strain MM 1 grown with MM. The
disappearance of (0) choline, (6) DM, and (D) MM (control) was followed by lH NMR
spectroscopy. Initial substrate concentrations of 6 g r1 and a protein concentration of 0.4 mg ml-1 were
used in all assays. Solid lines indicate initial slopes. Initial specific activities were 23 nmol rnin-1 mg-1
for choline, 48 nmol rnin-1 mg-1 for DM, and 324 nmol rnin-1 mg-1 for MM. The outlier is indicated in
brackets.
74
Microbial oxidation of MM
Products from choline. The lH NMR spectra of assays using the particulate fraction of strain
MM 1 and choline as substrate provided two product signals of methyl groups at 3.23 and
3.28 ppm additionally to the educt signal (Figure 4.4a). These products were identified as the
primary and secondary oxidation products betainealdehyde (hydrated form) and betaine,
respectively. The assignment of the NMR signals to the structures postulated was performed
based on the JH,l3C HSQC (correlation over one atom bond) and the lH, l3C HMBC spectra
(correlation over 2 or 3 atom bonds). All observed 2D correlation signals confirmed the
proposed structures. The chemical shifts 8eH) and 8(l3C) of both compounds are given in
Table 4.2 and the structures are shown in Figure 4.5a. The l4N NMR signal of
betainealdehyde was detected as broad shoulder of the main signal betaine. The definite
assignment of the signals was performed using a lH,l4N HMQC correlation. Whereas the l4N_
signal of choline was very narrow (~VIl2 =0.5 Hz), the signals of the products showed a clear
broadening of the lines (4.0 - 4.3 Hz).
Products from DM. The consumption of DM in particulate fraction enzyme assays was
monitored by lH NMR. The spectral region of the N-methyl resonances is shown in Figure
4.4b together with the postulated structure of the formed hemiacetal. This compound exists as
two enantiomers and only the structure of the 2R configuration is shown. Two lH NMR
signals of the diastereotopic methyl groups at 3.26 and 3.35 ppm were observed. All detected
lH, l3C correlation signals (HSQC and HMBC) confirmed the postulated structure
(Figure 4.4b). Chemical shift assignments are given in Table 4.3. The correlations H-(2)/C
(3), H-(5)/C-(1,4,6) and H-(6)/C-(1,4,5) observed in the HMBC spectrum demonstrated the
connectivity over the heteroatoms Nand O. With the lH,lH NOESY spectrum, the
neighbourhood of the anomeric proton H-(2) with the methyl group H-(6) was confirmed. A
8e4N) of -334.3 ppm was determined, corresponding to a deshielding of 6.2 ppm compared to
the educt signal. The line width of the product resonance (1.5 Hz) was only slightly enhanced
compared to the educt signal (0.5 Hz).
75
Chapter 4
Choline
Betainealdehyde
(a)
(b)
D
I I I I I8(1H) 3.40 3.35 3.30 3.25 3.20 ppm
Figure 4.4. lH NMR spectral region with signal assignements of the methyl groups of (a) choline and
(b) DM and their oxidation products. Initial substrate concentrations of 6 g r l, a protein concentration
of 0.5 mg ml- l and an incubation time of 12 hours were used for the enzyme assays.
Table 4.2. lH, 13C and l4N (with line width, ~Vl/2) chemical shifts of the products betaine and
betainealdehyde (hydrated form) obtained by degradation of choline in the particulate fraction of strain
MM 1. Chemical structures are shown in Figure 4.5a.
Betaine
Betainealdehyde
Position BeH) B(13e)[ppm] [ppm]
N-CH3 3.28 58.1N-CH2 4.02 69.9COOH 173.0N-CH3 3.23 58.9N-CH2 3.43 73.1CH-(OHh 5.56 89.6
Be4N) Llv 1/2
[ppm] [Hz]
-334.9 4.3
-335.2 ca.4
Table 4.3. lH and 13C chemical shifts of the product obtained from the degradation of DM in the
particulate fraction of strain MM 1. The positions of the carbon atoms in the molecule are shown in
Figure 4.5b.
Position of C NumberofH BeH) B(l3e)[ppm] [ppm]
1 2 3.31/3.56 67.32 1 5.40 92.43 2 4.03/4.34 60.64 2 3.52/3.52 64.75 3 3.35 58.76 3 3.26 57.6
76
Microbial oxidation of MM
(a) Structures of choline and its oxidation products
OH
IJOH IJ-OH
/N\ /N\
Choline Betainealdehyde(hydrated form)
Betaine
(b) Structures of DM and its oxidation product
DM
r 2 OH
N+""'''~s/~O
4
Oxidation product
(c) Structures of MM and its oxidation products
MMHOI+~OH
/N\/OH
7Ri6 2N+",,·,,~OH
S/ "y/304
trans-Products
IS 2
~ N+."",,~OHV "y/30
6 4
eis-Products
trans-/eis-1 :
trans-/eis-2:trans-/eis-3:
R = CH20H
R = CH(OH)2R=COOH
Figure 4.5. Chemical structures ofthe substrates (a) choline, (b) DM, (c) MM and their corresponding
oxidation products. Numbers indicate the position of carbon atoms.
77
Chapter 4
Table 4.4. IH and 13C chemical shifts of the products found by degradation of MM in the particulate
fraction of strain MM 1. The positions of the carbon atoms in the molecules are shown in Figure 4.5c.
Primary products: trans-l cis-l
Position of C NumberofH 8ctH) 8(13C) 8(IH) 8(13C)(ppm) (ppm) (ppm) (ppm)
1 2 3.39/3.69 66.6 3.39/3.61 66.7
2 5.44 92.4 5.41 92.3
3 2 4.05/4.40 60.0 4.09/4.31 61.0
4 2 3.57/3.69 64.3 3.59/3.65 64.2
5 3 3.42 55.5 3.32 54.1
6 2 3.65 71.2 3.78 72.1
7 2 4.12 59.4 4.12 59.6
Second. products: trans-2 cis-2Position of C NumberofH 8ctH) 8(13C) 8ctH) 8(l3C)
(ppm) (ppm) (ppm) (ppm)
1 2 * * * *
2 1 * * 5.40 *
3 2 * * * *
4 2 3.58/3.74 64.8 3.61/3.69 64.7
5 3 3.46 55.8 3.36 54.6
6 2 * * * *
7 1 5.64 89.1 5.66 89.3
Tert. products: trans-3 cis-3
Position of C NumberofH 8ctH) 8(13C) 8('H) 8(13C)(ppm) (ppm) (ppm) (ppm)
1 2 3.50/3.85 66.0 3.66/* 65.8
2 1 5.42 92.5 5.40 92.5
3 2 4.03 /4.40 60.1 4.09/4.32 60.8
4 2 3.61/3.87 63.7 3.66/3.76 63.9
5 3 3.52 55.5 3.41 54.6
6 2 4.02/* 69.7 4.09/4.24 70.0
7 172.9 173.4
* not assignable
78
Microbial oxidation of MM
Products from MM. The degradation of MM in PF enzyme assays (2 or 6 g r 1 MM)
provided at first an unidentifiable mixture of products. In the IH NMR spectrum, at least 7
different resonances of nitrogen-bound methyl groups were found in addition to the signal of
the educt MM (Figure 4.6b). As the assay proceeded, the signal of the methyl group of MM
(3.26 ppm) disappeared and the methyl resonances of the primary products at 3.32 and
3.42 ppm increased. With time, the primary product signals (assigned as trans-/eis-l in Figure
4.6) declined again while several secondary product peaks increased (Figure 4.6c). To
simplify the analysis of the spectra, MM was oxidised to the aldehyde by the Dess-Martin
method, since an oxidation was expected from the results obtained in anaerobic assays and
02-uptake experiments. The relevant section of a IH NMR spectrum of the synthesised
product, showing the methyl resonances, is depicted in Figure 4.6a. Additionally to the educt
signal (methyl groups of MM) two more singlets were detected at 3.32 and 3.42 ppm. They
correspond to the first evolving signals from the enzyme assay. Based on the IH NMR
spectrum two primary, diastereomeric oxidation products (trans-l and eis-I) were postulated
(Figure 4.5c). The stereochemical relations eis or trans are defined by the relative
configuration of the substituted carbon C-(6) to the OH group at the anomeric carbon C-(2).
As soon as one ethanol group of MM was oxidised an intramolecular reaction with a second
ethanol group lead to the cyclic hemiacetal (6-ring), similar to the cyclic structures found in
sugars (for glucose more than 99 % is usually present in the hemiacetal form; Koolman &
Rohm, 1998). Both diastereomeric products evolved in equal amounts.
In the l3C NMR spectrum 14 additional signals to the educt signals were detected. The IH, l3C
correlation over one bond (HSQC) showed distinct cross signals for 11 of these signals. Due
to extensive overlaps of the signals at about 4.1 ppm eH) and 59.8 ppm (l3c) with the strong
correlation signals of the residual educt, no unequivocal assignment was possible. Definite
assignment of the IH and l3C chemical shifts of trans-l (H-(3, 4, 7), C-(3, 7); Figure 4.5c) and
eis-l (H-(3, 7), C-(7); Figure 4.5c) was achieved by performing a HSQC-TOCSY experiment.
The IH, l3C HMBC spectrum showed all necessary correlation signals across the hetero atoms
Nand 0, which were essential to confirm the product structures. Moreover, the correlation
between the protons H-(2) and H-(5) (Figure 4.5c) found in the IH,lH NOESY spectrum
confirmed the relative configuration of eis-I. The corresponding protons of trans-l did not
show any cross signal.
79
Chapter 4
MM~ trans-/cis-1
811 trans-/cis-2
~ trans-/cis-3
(a)
?
?
(c)
(d)-A.. ~ -A.j
(e)
I3.50
I3.45
I3.40
I3.35
I3.30
I3.25 ppm
Figure 4.6. I H NMR spectra with assignments of the methyl groups of MM and its oxidation products
(a) from chemical oxidation of MM, (b) from an enzyme assay using the particulate fraction of MM
grown cells of strain MM 1 with 6 g r l of MM, 0.5 mg ml- I of protein, 12 hours of incubation, (c)
from a similar enzyme assay with 2 g r l of MM, 0.5 mg ml- I of protein and 5 hours of incubation, (d)
from the culture liquid of strain MM 1 grown in batch culture with an initial concentration of
350 mg r l of MM after 7 hours of incubation (exponential phase) and, (e) from the same culture after
12 hours (stationary phase) of incubation.
80
Microbial oxidation of MM
The 14N NMR signals were assigned via IH,14N HMQC correlation of the methyl protons at
3.32 and 3.42 ppm with the 14N resonances of both diastereomers at -329.0 ppm (trans-l) and
-327.8 ppm (eis-I) with line widths !1V1l2 of 2.5 and 2.3 Hz, respectively (Figure 4.7). The
average of 8ct4N) was -328.4 ppm and showed a shielding of 5.9 ppm relative to the 14N
signal of the educt. The chemical shifts of the primary oxidation products trans-l and cis-l
are listed in Table 4.4.
trans-2
~ ~trans-3
trans-1 G
-330.0
-329.0
Figure 4.7. IH,14N HMQC spectrum with assignment of signals to the oxidation products trans-/cis-l,
2 and 3 from an enzyme assay using the particulate fraction of MM-grown cells of strain MM 1. Initial
concentration of MM: 2 g r 1; protein concentration: 0.5 mg ml-1
; incubation time: 5 hours. MM was
completely converted to products under these conditions.
81
Chapter 4
In the IH NMR spectrum of the enzyme assay performed with MM at least 5 additional
signals with significant intensities were visible (Figure 4.6b,c), which belonged to the N
methyl groups of other products. In the IH,l3C-HMBC spectrum (correlation of the chemical
shifts over 2 - 3 atom bonds) explicit cross signals of protons in the region of 4.0 - 4.2 ppm
with carbon atoms at 172.9 and 173.4 ppm were detected. Each of these proton signals
correlated with three further carbon atoms providing very similar chemical shifts as found for
the products trans-I and eis-I. Based on these findings, two tertiary oxidation products trans
3 and eis-3 (Figure 4.5c) were postulated with the IH resonances of the methyl groups at 3.41
and 3.52 ppm (Figure 4.6b,c). The two compounds are cyclic hemiacetals as described for the
primary products with the third ethanol group oxidised twice to the corresponding carboxylic
acid. The chemical shifts 8eH) and 8(l3C) of both cis- and trans-3 are given in Table 4.4.
Again, both products were found in equimolar amounts.
The IH, l3C HMBC correlation of the methyl protons in position 5 with the carbon atoms 1,4
and 6 (Figure 4.5c) as well as the correlations found with the IH,l3C HSQC spectrum
confirmed the structural elements linked to the nitrogen atoms. The IH, l3C HSQC-TOCSY
experiment provided the assignment of the IH and l3C chemical shifts at positions 2 and 3
(Table 4.4). The IH,lH NOESY spectrum showed unequivocally that the signal of the methyl
group at 3.41 ppm correlated with H-(2), therefore confirming the structure eis-3. The
structure of the trans-3 product was confirmed via the correlation between the protons H-(6)
and H-(2) (Figure 4.5c). In the ID 14N NMR spectra the signals of the different products were
not resolved clearly due to strong overlapping, and in samples with low product
concentrations these signals were hardly detectable. Therefore, the assignment of the signals
to structures trans-3 and eis-3 was performed via the IH,14N HMQC correlation (8(14N):
trans-3: -330.5 ppm; eis-3: -329.9 ppm; Figure 4.7).
Additionally to the previously assigned IH NMR signals, two further methyl singlets were
detected at 3.36 and 3.46 ppm (Figure 4.6b,c), which were supposed to belong to the cyclic
products trans-2 and eis-2 (hydrate form of the aldehyde, Figure 4.5c). Because products with
a higher oxidation state (carboxylic acid) were found and confirmed, one would also expect
these intermediates.
Since most resonances of the IH and l3C signals probably lay beyond the signals trans-/cis-I,
trans-/eis-3 and MM (Figure 4.6b,c), no final prove of eis-2 and trans-2 was possible without
82
Microbial oxidation of MM
purification of individual products from the mixture. However, several indications were found
for their existence and the chemical shifts 8eH) and 8(13C) giving evidence to the proposed
trans-2 and eis-2 structures are listed in Table 4.4.
The IH,13C HMBC correlations of the methyl protons H-(5) (Figure 4.5c) provided cross
signals to the carbon atoms at 64.7 and 64.8 ppm (probably C-(4)). The correlation signals to
C-(I) and C-(6), supporting these structures were not detected unequivocally. Additional hints
for the expected structures were observed in the IH,lH NOESY spectrum: trans-2 showed a
correlation signal of H-(5) with H-(7), whereas for cis-2 a correlation of H-(5) with H-(7) as
well as with H-(2) was detected. Based on this steric interaction the relative configuration of
cis-2 at the nitrogen atom was deduced. The corresponding 14N chemical shifts of -330.3 ppm
(trans-2) and -329.7 ppm (eis-2) were observed (Figure 4.7).
All coincident pairs of the described diastereomers showed a stronger shielding of the
nitrogen nucleus in the eis-configuration compared to the isomers trans. It was not possible to
identify the structure of the product with its methyl singlet at 3.34 ppm (Figure 4.6b,c).
During batch growth of strain MM 1 the substrate MM continuously decreased in the
exponential phase until it had completely disappeared when reaching the stationary phase
after about 11 hours (Chapter 3). However, at the end of such batch cultures, high residual
concentrations of carbon and nitrogen were left in the culture broth. Therefore, supematants
of batch cultures in the exponential and stationary phase were investigated by IH NMR
spectroscopy with respect to the presence of excreted metabolites. Such products were found
in considerable amounts in the culture liquid and the majority of them were identified as
tertiary oxidation products eis-/trans-3 (Figure 4.6d and e). Based on the measured residual
carbon and organic nitrogen concentrations (Chapter 3) these products accumulated during the
exponential growth phase to a maximum final concentration of 30 mole-% of the initially
provided MM. In the exponential phase, transient accumulation of low concentrations of the
primary oxidation products cis-/trans-l (Figure 4.6d) were detected as well. However, they
disappeared again in the late exponential phase (Figure 4.6e).
83
Chapter 4
Expression of enzymes in the particulate fraction of MM 1. The expression of the enzymes
responsible for the degradation of MM, DM or choline was studied with the help of enzyme
assays employing the particulate fraction of MM 1 cells grown either with MM (control),
choline or acetate. Acetate was chosen as a substrate because it does not contain nitrogen and
is catabolised via a different metabolic pathway compared to that known for choline. For the
growth with acetate N~Cl was used as nitrogen source. The experiments were performed
with 2 g r l of the carbon substrate and a protein concentration of 0.5 mg mr l of the protein
fraction. After 5 hours of incubation the reaction was stopped and the assays were analysed by
I H NMR spectroscopy. In all assays transformation of MM, DM and choline to the
corresponding products was observed as described above. MM always was converted to the
highest extent, followed by DM and choline (Figure 4.8).
Strain MM 1 grown with:
Choline
100 ~,="....---_......!-.,--!----~.,--!------L....,
~ 80E::JC/)c8 60Q)
co....en 40.0::JC/)
<5~ 20o
oMM DM Ch MM DM Ch MMDM Ch
Substrates tested in enzyme assays
Figure 4.8. Substrate specificitiy of the particulate fraction of strain MM 1 grown with different
substrates. The bars indicate the percentage of consumed substrate after 5 hours of incubation (single
experiments). The enzyme assays were performed with initial concentrations of MM, DM and choline
(Ch) of 2 g rI, each, and protein concentrations of 0.5 g rl.
84
Microbial oxidation of MM
DISCUSSION
All detected enzyme activities were associated with the membrane of strain MM 1 making
purification of the enzymes difficult or even impossible. Therefore, enzyme assays for the
investigations of the catabolic pathways were performed with the particulate fraction of strain
MM 1.
The responsible enzyme for the initial degradation of methyl-triethanol-ammonium (MM)
most probably belongs to the membrane-associated oxidoreductase group of enzymes. Under
aerobic conditions, molecular oxygen acted as electron acceptor. Based on the performed
experiments, the electrons derived from MM were transported to molecular oxygen rather via
the respiration chain present in the particulate fraction than directly by the "MM
oxidoreductase". However, the unequivocal elucidation of the path of the electrons requires
additional investigations.
The proposed degradation pathway of MM is depicted in Figure 4.9. The products trans-lcis-l
and trans-lcis-2 are expected to undergo further degradation. Obviously, the continuing
degradation of the ring structure is not mediated by the "MM-oxidoreductase" since the cyclic
hemiacetals prevented further oxidation of the ethanol and acetaldehyde groups involved in
the cycle. The products trans-Icis-3, with the third ethanol group oxidised to the carboxylic
acids, are most likely dead-end metabolites, because considerable amounts of these
compounds were released into the culture broth and remained untouched in batch cultures of
strain MM 1. Consequently, one has to take into account that these metabolites may
accumulate in the environment. To investigate the possible presence of these compounds in
the environment, extensive studies in complex environmental systems, i. e. river water or
sewage treatment sludge would be required turning one's attention to the analysis of the
described metabolites. Detection of such metabolites in the environment would greatly
influence the design of new similar compounds.
The cyclic oxidation products (trans-Icis-l, 2 and 3) of MM each existed as 1: 1 mixture of the
two diastereomers. This fact may be interpreted in two ways: Either the enzyme mediating the
first oxidation step to the cyclic hemiacetals trans-Icis-l was not a diastereoselective reaction
(as would be expected) and/or this cyclic hemiacetal underwent a rearrangement with the
remaining ethanol group in the aqueous environment.
85
Chapter 4
Regarding the substrate specificity of the particulate fraction, the oxidation of DM and
choline proceeded under the same conditions as that found for MM, whereas TM remained
untouched. Additionally, the MM-, DM- and choline-consuming activity was independent on
the growth substrate used for strain MM 1. This suggests that one and the same constitutively
expressed enzyme catalyses these reactions. Since strain MM 1 was able to grow with choline
as well and the ability to oxidise choline is widespread amongst microorganisms (Kortstee,
1970) the enzyme described even may be a choline oxidoreductase with a broad substrate
specificity. Based on these findings, the following conclusions can be drawn with respect to
the described oxidoreductase. Free ethanol groups, as present in MM, DM and choline, are
essential to undergo oxidation in the active enzyme and/or the 2,3-dihydroxy-propyl part of
MM
trans-/cis-1
trans-/cis-2
~2e-+2H'OH
\,••/\OH ~/V-
0
~
H20-.,J Il\.... Further breakdownOH 20-+2/----------
HO~ OHN+''''''''.!\
/V- 0
~2e-+2H'trans-/cis-3
o
HO~ OH
N+''''''''.!\/V- 0
Excretion asdead-end metabolite
Figure 4.9. Proposed pathway for the initial degradation steps of MM in strain MM 1. Only trans
structures with the absolute configuration 2R, NS are shown.
86
Microbial oxidation of MM
TM prevented the molecule to fit into the active site. Obviously, the presence of a quaternary
nitrogen was not sufficient to allow enzymatic oxidation of the hydroxyl groups. Since DM
was oxidised in the particulate fraction but did not serve as a growth substrate, strain MM 1
was probably hampered in the transport of this compound into the cell.
Membrane-associated oxidoreductases consuming choline with similar properties as described
here were also detected in Pseudomonas aeruginosa and Escherichia coli (Bater & Venables,
1977; Lamark et al., 1991; Nagasawa et al., 1976; Russell & Scopes, 1994). Surprisingly,
these oxidoreductases (choline dehydrogenases) specifically oxidised choline to
betainealdehyde only, without further oxidation to betaine. Enzymes mediating both oxidation
steps were characterised by Ohta-Fukuyama et al. (1980) and Ikuta et al. (1977) from
Alcaligenes spec. and Arthrobacter globiformis, respectively. However, these enzymes were
soluble oxidases and both of them produced HzOz.
The particulate fraction described in this work exhibited a specific acitivity with choline in the
range of about 25 nmol min- I mg- I. This corresponds well to the specific activity range for
choline dehydrogenases in cell-free extracts and particulate fractions of
7 -78 nmol min- I mg- I reported in the literature (Boncompagni et al., 1999; Nagasawa et al.,
1976; Pocard et al., 1997).
The presented study suggests that the oxidation of MM might be linked to the oxidation of
choline. This raises the question, whether or not the degradation of quaternary ammonium
alcohols in general is related to the choline degradation pathway or whether different
strategies are responsible for the degradation of different quaternary ammonium alcohols. To
answer this question further investigations at the enzyme level using different microorganisms
able to grow with TM and DM are needed.
87
eite Leer /Blank leaf
Microbial degradation of TM
5. Microbial degradation of 2,3-dihydroxypropyl-trimethyl-ammonium
ABSTRACT
2,3-dihydroxypropyl-trimethyl-ammonium (TM) originates from the hydrolysis of the parent
esterquat surfactant, which is widely used as softener in fabric care. Judged on OECD
standard biodegradation tests simulating complex biological systems, TM is supposed to be
biologically degraded to biomass and CO2 when reaching the environment. However, the
degradation mechanisms of TM have not been elucidated so far. The initial step of breakdown
of TM was investigated in cell-free extracts of strain P. putida TM 1, a bacterium able to
grow with TM as sole source of carbon, energy and nitrogen and to degrade this compound to
completion. TM-consuming activity was located in the particulate fraction of P. putida TM 1.
Trimethylamine was split from TM without the addition of any cofactors and independent of
the presence of oxygen. Therefore, the responsible enzyme was supposed to be a membrane
associated lyase. Unfortunately, the structures of the remaining products derived from the
propyl moiety have not been identified so far, although their presence was detected by
different analytical methods. "TM-Iyase" appears to be an inducible enzyme because the
ability to catabolise TM was solely observed in the particulate fraction of TM-grown cells and
no activity was found in the particulate fraction of cells grown with acetate or choline. Except
for choline, none of the tested structurally related compounds such as dimethyl-diethanol
ammonium, methyl-triethanol-ammonium (both also used as head groups in esterquat
surfactants), betaine, D- or L-carnitine, were converted in the particulate fraction. However,
choline was transformed by a different mechanism, namely by an oxygen-dependent reaction.
The products from choline were identified as the primary and secondary oxidation products
betainealdehyde and betaine. All this indicates that the initial TM degradation is a highly
specific reaction and that there is no involvement of enzymes responsible for choline
catabolism.
89
Chapter 5
INTRODUCTION
The quaternary ammonium alcohol 2,3-dihydroxypropyl-trimethyl-ammonium (TM),
dimethyl-diethanol-ammonium (DM) and methyl-triethanol-ammonium (MM) belong to the
three mainly used head groups in cationic fabric softeners (Krueger et al., 1998). The parent
esterquat surfactants consist of the quaternary ammonium alcohols (QAAs) esterified at two
alcohol groups with long chain fatty acids deriving from tallow. Since the esterquat
surfactants are produced in considerable amounts worldwide (probably more than 100000
tons; Krueger et al., 1998), they have been investigated extensively in DECD test procedures
and by monitoring their environmental concentrations. When reaching surface water or
sewage treatment plants they hydrolyse rapidly, abiotically and/or biocatalysed, to the fatty
acids and the corresponding QAAs. Based on DECD biodegradation tests, both, the parent
compounds as well as their products are judged as readily and completely biodegradable
(Giolando et al., 1995; Krueger et al., 1998; Matthijs et al., 1995; Puchta et al., 1993; Simms
et al., 1992; Waters et al., 1991; Waters et al., 2000). Whereas the fatty acids are expected to
degrade via the common fatty acid metabolism (l3-oxidation), the degradation mechanisms for
the QAAs are not known to date. Details are limited to some DECD die-away tests, which
provided different degradation rates and patterns for the three QAAs (Hales, 1998).
Considering the large quantities used worldwide and the development of new similar head
groups with improved environmental properties, it is important to elucidate the strategies and
mechanisms involved in their biodegradation. Therefore, we isolated microorganisms able to
grow with these QAAs as a basis for further biochemical investigations.
Using the quaternary ammonium alcohol TM as a sole source of carbon, energy and nitrogen,
a Pseudomonas putida strain, referred to as P. putida TM 1, was isolated and described in
detail by Kaech & Egli (2001). In this study, the initial degradation mechanism of TM in cell
free extract of this isolate was investigated.
MATERIALS AND METHODS
Chemicals. The quaternary ammonium alcohols (QAAs) (±)-2,3-dihydroxypropyl-trimethyl
ammonium (TM, racemic mixture 1:1), dimethyl-diethanol-ammonium (DM) and methyl-
90
Microbial degradation of TM
triethanol-ammonium (MM) were provided by Unilever (SEAC Safety and Environmental
Assessment Center, Unilever Research, Port Sunlight, UK) as the iodine salts. 14C-labelled
TM was supplied as stock solutions in methanol. The activity of 1 /11 of 14C-methyl-labelled
TM and 14C-propyl-labelled TM (label at propyl carbon in position 3) was 36000 dpm and
85000 dpm, respectively. All other chemicals were purchased from Fluka, Buchs,
Switzerland, unless indicated otherwise.
Bacterial strains and storage. Experiments were performed with cells and cell-free extracts
of strain P. putida TM 1, which was isolated with TM as the sole source of carbon, energy
and nitrogen (Kaech & Egli, 2001). In some experiments, also the particulate fraction of
isolate DM 2 was used, a strain isolated originally with DM, but also able to grow with TM
and using both substrates as the sole source of carbon, energy and nitrogen. The 16S-rDNA
sequence of isolate DM 2 indicated identities to Zoogloea ramigera Itzigsohn 1868AL of 98 %
(best hit, total 1448 base pairs, Chapter 3). For short-term storage all strains were plated on
lO-fold diluted tryptic soy agar or on agar plates containing SM and a selective carbon source.
For long-term preservation all strains were suspended in 30 % glycerol and stored at -80 QC.
Cultivation of microorganisms. Cultivation of the microorganisms was performed as
described by Kaech & Egli (2001).
Carbon balances using 14C-Iabelled TM. Carbon balances for P. putida TM 1 cultures were
performed as follows. During the exponential phase of a culture growing with TM (conditions
and medium as described by Kaech & Egli, 2001), 15 ml were transferred into a 50 ml
Erlenmeyer flask and incubated at room temperature. Aeration of the culture was achieved
using a magnetic stirrer. After the addition of 14C-labelled TM, the flask was sealed with a
silicon stopper. The culture was bubbled with air and the air outlet was connected to two
vessels (7 ml) in sequence containing each 5 ml of Carbo Sorb (Packard Bioscience
Company, Groningen, The Netherlands) to absorbe the produced CO2. Samples of the culture
and the C02-absorbing liquids were taken immediately after addition of the 14C-Iabelled TM
and after incubation for three hours. 14C-Iabel incorporated into biomass was determined by
filtering 2 ml of the culture liquid through a 0.45 /1m cellulose nitrate membrane filter
91
Chapter 5
(Millipore, Volketswil, Switzerland) using a vaccum pump. Filters and collected cells were
washed with distilled water containing 15 mM cold TM and thereafter dissolved in 3 ml of
Filtercount scintillation cocktail (Packard Bioscience Company). Samples of the filtrates and
the culture broth were transferred to scintillation vials containing 4 ml of Instagel plus
(Packard Bioscience Company) and aliquots of the CO2-absorbing vessels were put into
scintillation vials containing 5 ml of Permafluor E+ scintillation cocktail (Packard Bioscience
Company). The radioactivity obtained in the different fractions was then determined in a
liquid scintillation analyser (Tri-Carb 2200 CA, Packard, USA).
Preparation of cell-free extracts. The preparation of the cell-free extracts of P. putida TM 1
and the isolate DM 2 were performed as described in Chapter 4.
Enzyme activity assays. Enzyme activity assays were carried out in GC-Vials (Supelco Inc.,
Bellefonte PA, USA) of 5 to 20 ml volume. Protein solutions were stirred with a magnetic
stirrer and incubated at a pH of 7 in a 30 QC water bath. Using NMR spectroscopy as the
analytical method, the samples were either diluted with D20 (Aldrich Chemical Company
Inc., Milwaukee Wi, USA) or PB in D20 was used in the assay. The concentration of protein
used in assays was always in the range of 0.4 to 1.7 mg mrl. To stop the reaction, samples
removed from the assay were either heated for 1 min in a water bath of 95 QC or HCI 1 M was
added (l0 % of sample volume). Before analysis, acidified samples were neutralised with
NaOH 1 M (10 % of sample volume) and the pH was controlled with test strips (Tritest
pH 1 - 11, Macherey-Nagel, Dtiren, Germany). The stop procedure performed with acid did
not change the chemical composition of the samples as observed by lH NMR analysis. To
exclude oxygen, experiments were carried out in sealed vessels with a continuous nitrogen
flow through the headspace. All chemicals used in these experiments, except for the protein
stock solution, were bubbled with nitrogen for 10 min before adding them to the assay. NAD+
and phenazine methosulfate (PMS) were tested as electron acceptors. They were prepared as
stock solutions and the final concentration used in individual assays is indicated at the
appropriate place. For control experiments, the protein solution was heated in a water bath
(-95 QC) for 1 min prior to the addition of the different substrates. Samples removed from
assays were stored at -20 QC prior to analysis. Enzyme assays were also carried out directly in
92
Microbial degradation of TM
NMR-tubes (diameter: 5 mm) in the NMR spectrometer and IH NMR spectra were acquired
as a function of reaction time. Between the acquisition of lH NMR spectra, the tube was
rotated to mix the protein solution, since the PF sank to the bottom of the tube without
agitation.
Assays to check the presence of TMA dehydrogenase or monooxygenase in CFE and SF were
performed as described by Colby & Zatman (1972). Samples were analysed by UV-VIS
spectrophotometry and by IH NMR spectroscopy.
To check the presence of formaldehyde dehydrogenase, the following enzyme assays were
performed. To 1 ml of 100 mM sodium phosphate buffer (pH 8) in a quartz cuvette, NAD+
and glutathion were added to a final concentration of 2 mM and 3 mM, respectively.
Subsequently, the protein fraction to be tested (0.14 mg mrl final concentration in the assay)
and formaldehyde (1 mM final concentration) were added, and the absorption of NADH was
monitored at 340 nm in a spectrophotometer. Formate dehydrogenase was measured by a
similar procedure. However, the pH of the phosphate buffer was 7.5, glutathion addition was
omitted and sodium formate (pH =7) was used as substrate (1.7 mM final concentration). To
calculate specific activity the initial slopes and an extinction coefficient of 6200 M'I cm'l for
NADH was used. All assays were done at 30°C.
Specific activities are given in nmol of converted substrate per min and per mg of protein
(nmol min'l mg'I).
To determine protein concentration the Bio-Rad enzyme assay was used according to the
manual of the manufacturer (Bio-Rad Laboratories GmbH, Munich, Germany). BSA was
used as standard (A-7906, Sigma-Aldrich Chemie GmbH, Steinheim, Germany).
Filtration. Cell constituents and proteins were removed from samples by filtration. For small
volumes (0.5 ml), the samples were centrifuged at 13000*g in a microcentrifuge at room
temperature using Ultrafree centrifugal filter devices (Biomax, cut-off 5000 Da, Millipore
Corporation, Massachusetts, USA). For volumes of 10 ml, larger Centriplus centrifugal filter
devices (YM-3, cut-off 3000 Da, Millipore) were used, centrifuging at 3000*g at room
temperature for several hours. Since the filter membranes were humidified with glycerol, all
filter units were prewashed several times with distilled water. As controls samples of the last
93
Chapter 5
washing step were analysed with the same method as the samples obtained from the enzyme
assays.
Thin layer chromatography (TLC). Samples from enzyme assays performed with 14C_
labelled TM as a substrate were separated by thin layer chromatography (TLC) using silica
gel coated aluminium plates (DC-Alufolien Kieselgel 60 F254, 20 x 20, Merck KgaA,
Darmstadt, Germany). Methanol: ethylacetate (1:1) or acetone: NaCI 1 M (1:1) were used as
eluents. Appropriate amounts (5 - 20 J.tl) of the samples were either directly applied onto the
TLC plates or proteins were removed previously by filtration. After chromatographing the
plates in a closed and eluent saturated TLC chamber (incubation time about 1 hour), they
were dried at room temperature and exposed to a Biomax film (Eastmann Kodak Company,
Rochester, New York, USA) for 4 to 7 days at -80 QC using a cassette with an intensifying
screen. Afterwards, films were developed in an automatic Kodak developer.
Extractions. Metabolites from enzyme assay samples were extracted with diethylether,
methyl acetate, and ethyl acetate. The volume of the solvent used was twice that of the
sample. When two extraction steps were performed the organic phases were combined for
analysis. From extractions carried out with 14C-Iabelled TM (propyl- or methyl-labelled) the
water phase (before and after extraction), the interphase and the organic phase were analysed
by scintillation counting. The interphase was analysed as well because a considerable amount
of TM aggregated in this layer. Aliquots of the samples and extracts were added to 3 ml of
Insta-Gel Plus scintillation cocktail (Packard BioScience Company, Groningen, The
Netherlands) and the radioactivity was measured in a liquid scintillation analyser (Tri-Carb
2200 CA, Packard, USA). Based on this, the extraction efficiency was calculated. Control
experiments were performed with inactivated protein solutions.
Desalting. Desalting was performed to remove the sodium phosphate buffer and other
charged components from the sample. A desalting chamber of 2.5 ml volume with a cation
and an anion-selective membrane (Berghof, Enningen, Germany) was used. A cooled (4 QC),
circulating NaCI solution (0.5 g r1) served as electrolyte. Power (final voltage: 500 V) was
applied to the chamber until the power supply indicated no current anymore.
94
Microbial degradation of TM
Lyophilisation. Freeze-drying was applied to concentrate product components from the
different fractions (original samples, protein-free samples and/or desalted samples). The
solutions were frozen at -80 QC and subsequently dried in a Lyovac GT2 (Leybold AG, Koln,
Germany).
Nuclear magnetic resonance spectroscopy (NMR). IH NMR spectroscopy was used as the
analytical method to follow enzymatic activity and to determine the reaction products.
IH NMR spectroscopy was performed as described in Chapter 4. The chemical shifts of TM
and choline relative to 3-trimethylsilyl-tetradeutero-propionate are listed in Table 4.1,
Chapter 4.
Ion-pair chromatography. Additionally to IH NMR spectroscopy, TM was measured by
ion-pair chromatography as described by Kaech & Egli (2001).
Spectrophotometric determination of TMA. TMA concentrations in batch cultures were
determined by the spectrophotometric method as described by Shen (1988).
GC-MS (gas chromatography coupled with mass spectrometry). For the analysis of
expected products GC-MS was performed using a gas-chromatograph Fisons 8000 with PTV
injector OPTIC 2 (ATAS, Veldhoven, The Netherlands) and a high resolution mass
spectrometer (Autospec S, Micromass, Manchester, UK). GC-parameters: 1 J.tl injection
volume; column: XTI-5, 30 m, 0.25 mm internal diameter, 0.25 J.tm film thickness (Restek
Corp., Bellefonte, PA, USA) with a 2 m pre-column Siltek of 0.53 mm internal diameter
(Restek); oven temperature program: 50 QC for 2 min, gradient 9 QC min'I to 320 QC, followed
by 320 QC for 5 min, giving a total run time of 37 min. Injector parameters: Equilibration
time, 0.00 min; initial temperature, 270 QC; ramp rate, 0 QC min,I; final temperature, 270 QC;
splitless time, 1.5 min; transfer pressure, 0.85 bar; transfer time, 1.5 min; initial run pressure,
0.85 bar; final run pressure, 2.0 bar; end time, 55 min; purge gas flow, 2 ml min-I; split gas
flow, 50 ml min'I. MS-parameter: Ionisation-mode, +EI; acceleration voltage, 8000 V;
transferlines, 300 QC; source temperature, 280 QC; electron energy, 70 eV; trap current,
95
Chapter 5
500/lA; detector, 270 V; calibration, perfluorokerosene low boiling; resolution, 1000 (5 %
Valley); mass-range, 60-500 Da; scan time, 0.9 s; delay time, 0.5 s.
The ethylacetate phase of extracted samples was directly injected into the system. Freeze
dried samples of enzyme assays were dissolved in a mixture of ethylacetate (200 ILl) and
methanol (300 ILl) before injection. For derivatisation of formed metabolites a mixture of
N-methyl-N-trimethylsilyltrifluoroacetamid, l,4-dithioerythritol (Merck, Darmstadt,
Germany) and l-(trimethylsilyl)imidazole v:w:v = 1000:2:2 was used. To perform the
reaction 50 ILl of the reagent was added to the freeze-dried samples and incubated 30 minutes
at 60 cc. Afterwards, the derivatisation reagent was evaporated with nitrogen and 200 ILl of
hexane was added to dissolve the products that might have been formed.
ESI-MS (Electrospray ionisation mass spectrometry). Samples of enzyme assays were also
analysed by direct infusion ESI-MS (TSQ Quantum, Finnigan, CA, USA). ESI parameters
were: Infusion rate, 40 ILl min- I; spray voltage, 4000 V (ESI+), 2500 V (ESr); sheath gas
pressure, 39 psi; auxiliary gas pressure, 5 psi; capillary temperature, 350 cC; tube lens offset,
61 V. Scan parameters: Mass range, 50 - 500 Da; scan time, 0.75 s; peak width, 1.00 Da.
RESULTS
Conversion of TM. Disappearance of TM was observed in cell-free extract and in the
particulate fraction of P. putida TM 1 cells grown with TM as sole source of carbon, energy
and nitrogen. In contrast, TM was neither consumed in the soluble fraction independent of the
presence or absence of 2 mM NAD+, nor in control experiments with heat-inactivated
extracts, indicating that TM-transforming enzymes in P. putida TM 1 are membrane
associated. Addition of cofactors or cosubstrates was not required for TM-consuming activity
in the PP. Even under anoxic conditions TM was converted. This points to a lyase
mechanism. The addition of PMS (2 mM) did not show an effect on the reaction and NAD+
(2 mM) even inhibited consumption of TM. In enzyme assays with initial TM concentrations
of 50 mg r l up to 6 g r l the reaction always levelled of when 29 ± 5 % (6 values) of TM had
been consumed, independent on the protein concentration used (between 0.4 and 1.7 mg mr l).
The extent of - 30 % conversion was not caused by enantioselectivity of the enzyme system,
96
Microbial degradation of TM
since a 1 : 1 racemic mixture of TM was used and therefore would yield 50 % of TM being
consumed. When initial concentrations of 10 and 15 g r 1 of TM were used, the amount of
consumed TM was reduced to 15 % and 10 %, respectively, even after extended incubation
times of up to 24 hours. With initial TM concentrations of 50 to 600 mg r1 the final level of
TM consumption was reached after approximately 1.5 hours (Figure 5.1). The initial specific
activity in enzyme assays with 1.5 mM of TM was 4 - 5 nmol min-1 mg-1 only, based on IH
NMR spectroscopy (Figure 5.1) and ion-pair chromatography data.
0"..... 0
o 08 0 0
/ t ~ ~
--
2.0
1.5
~.s« 1.0:2I-
~I-
0.5
0.0o 2
Time [h)3 4
Figure 5.1. Disappearance of TM (0) and appearance of TMA ( ... ) in the particulate fraction of
P. putida TM 1 grown with TM as sole source of carbon, energy and nitrogen, as determined by
IH NMR spectroscopy. Values are based on the absolute integrals of the CH3-signals in the IH NMR
spectra. Initial TM concentration: 1.5 mM. Protein concentration used in the assay: 1.7 mg ml-1• The
specific activity based on the initial slope was 4.4 nmol min-1 mg-1 (solid line). Note: The final TM
level was reached already after 80 min of incubation with only 20 % of TM being converted. Since the
experiment was carried out directly in an NMR tube in the NMR spectrometer, the mixing most
probably was insufficient (only rotation of the tube between data acquisition procedures) and,
therefore, the extent of TM conversion of roughly 30 % typically observed in experiments with intense
stirring was not reached.
97
Chapter 5
Trimethylamine. In IH NMR spectra recorded during enzyme assays with PF and CFE,
TMA was detected as a product of the consumption of TM and TMA was formed
stoechiometrically to the disappearance of TM (Figure 5.1). Since TMA possesses nine
equivalent protons (as does TM), the IH NMR spectra provided a clear an intensive singlet
(Figure 5.2). Because only approximately 30 % of the originally present TM was converted,
TMA and the suspected product of the reaction (i. e. glycidol, see section "Unidentified
products") were investigated with respect to their influence on the extent of conversion. Either
TMA (1.5 mM) or glycidol (1.5 mM) were added to the assay before starting the reaction with
TM (1.5 mM). Astonishingly, no product inhibition or change in the percentage of consumed
TM was detected. Also readdition of the same amount of PF to an assay with 1.5 mM TM
after the reaction had levelled off lead to no further degradation of the residual TM.
(a)/ OH
" ~N~OH""-/ TM
I3.30
I3.20
I3.10
I3.00
I I2.90 8(1H) ppm
Figure 5.2. IH NMR spectra ofthe enzyme assay shown in Figure 5.1 after (a) 2 min and (b) 80 min
of incubation. Initial TM concentration: 1.5 mM. Protein concentration of the particulate fraction of
P. putida TM 1 grown with TM as sole source of carbon, energy and nitrogen: 1.7 mg ml-1. Only the
section of the CH3-group signals is depicted. Arrows show the assignment of the signals to the
corresponding group in TM and TMA, respectively.
98
Microbial degradation of TM
Based on these observations above, several experiments were carried out to examine whether
or not in the PF of P. putida TM 1 formation of TMA from TM was an artifact or a
physiological property. The degradation of TM was investigated in cell-free extracts of
P. putida TM 1 grown with either acetate or choline but neither in the CFE nor in the PF of
these cells TMA was formed in enzyme assays using 15 mM of TM. This suggests that the
enzymes acting on TM are inducible and that the reaction is not an artifact of the PF itself. PF
also was prepared from isolate DM 2, belonging to a different bacterial genus, when grown
with TM as sole source of carbon, energy and nitrogen (Chapter 3). In an enzyme assay using
this PF (l mg mr l of protein) and a TM concentration of 1.5 mM of TM, TMA was also
formed and NAD+ inhibited the reaction as well.
TMA formation was also investigated with growing cells of P. putida TM 1. Additional TM
(30 mM) was either pulsed to a culture grown with 3.7 mM of TM in the stationary phase, or
cells from the exponential growth phase of a batch culture growing with 30 mM of TM were
harvested by centrifugation and suspended in new medium containing 30 mM of TM. In both
experiments the time course of TMA concentration in the medium as well as the optical
density were monitored. The results obtained when resuspending the cells in fresh medium
are shown in Figure 5.3.
1.0 0.35
0.9 0.30
0.250.8
E 0.20 ~c: .sco<t 0.7l!l «Cl
0 0.15 ~I-
0.60.10
0.5 0.05
0.4 0.00
0 2 3 4 5 6 7 8 9Time [h]
Figure 5.3. Time course of the optical density ( 0 ) and the TMA concentration ( ... ) in a batch culture
of P. putida TM 1 after resuspending cells harvested from the exponential phase of a batch culture in
fresh medium. The initial TM concentration was 30 mM in both cultures.
99
Chapter 5
In both experiments excretion of TMA into the culture medium was detected, confirming that
TMA formation from TM did also occur in living cells of P. putida TM 1. Although, the
maximum concentration of TMA excreted into the medium was very low, ~ 0.1 mM in pulse
experiments, and ~ 0.3 mM after resuspension (Figure 5.3), the formation of TMA was
unambiguous. Undisturbed batch cultures, even with NH4CI (800 mg r1) as additional
nitrogen source, and chemostat cultures of P. putida TM 1 growing with TM never contained
detectable levels of TMA.
If TMA was a metabolite in TM degradation, one would expect that TMA can be used as a
nitrogen source by the organism. This was tested in batch and continuous cultures of P. putida
TM 1 with either acetate or ethanol as additional carbon source and TMA as the sole source of
nitrogen (both, acetate and ethanol can be used as carbon sources for growth by
P. putida TM 1; Kaech & Egli, 2001). For this, 10 % (v/v) of a P. putida TM 1 batch culture
(3.7 mM TM) from the exponential phase was transferred into a shake flask containing fresh
medium with 17 mM of acetate and 3 mM of TMA at a pH of 9. A batch with 17 mM of
acetate only was used as a control. A pH of 9 was chosen to increase the concentration of non
protonated TMA available (~ 14 % of total TMA is present in the neutral form at pH = 9) that
is expected to enter the cells by simple diffusion. No difference was found between the batch
with TMA as nitrogen source and the control, strongly suggesting that TMA was not used as
nitrogen source. A slight growth was observed in both cases and may be caused by the
residual nitrogen transferred with the inoculum (cells not washed) and/or by an accumulation
of surplus carbon as storage material.
The utilisation of TMA as a nitrogen source also was investigated by pulsing TMA to a
nitrogen-limited continuous culture of P. putida TM 1 at a dilution rate of 0.1 h-1 and fed with
2.6 mM (350 mg r 1) TM and 17 mM (l g r 1
) ethanol at 30 QC and a pH of 8 (~2 % of total
TMA present in the non-protonated form). TM was provided as a substrate to guarantee the
expression of all enzymes required for TM utilisation. Ethanol as an additional carbon source
was used to ensure the assimilation of surplus nitrogen from TM, usually released as
ammonium. The ratio of TM : ethanol was set such that neither residual TM nor excreted
NH/, but residual carbon (~ 850 mg r1) was present. When the steady-state was reached,
TMA was pulsed to the culture to a final concentration of 0.85 mM and the biomass (optical
density), dissolved organic carbon and ammonium were measured before the pulse and during
100
Microbial degradation of TM
8 hours following the pulse. The ammonium concentration remained below the detection
limit, additional excess ethanol was not utilised (this would have resulted in a decrease in
DOC) and the biomass concentration remained constant. This clearly demonstrates that also
under these conditions TMA was not utilised as a nitrogen source. As a control,
monomethylamine (MMA) was pulsed to the same culture (final concentration 1.4 mM),
since MMA was known to be used by P. putida TM 1 as a nitrogen source in batch cultures.
Four hours after the pulse, the biomass started to increase dramatically, confirming the
utilisation of MMA as a nitrogen source and the validity of the experimental setup.
Since TMA was formed in the PF, TMA-degrading activity was expected to be present in
CFE or SF of P. putida TM 1. Therefore, the assays for TMA dehydrogenase and TMA
monooxygenase originally developed for methylotrophs (Colby & Zatman, 1972) were used
to detect activity of TMA-utilising enzymes. However, no activity of neither enzyme was
found in CFE or SF ofTM-grown P. putida TM 1.
Unidentified products. Since TMA was removed enzymatically from TM in the PF of
P. putida TM 1 the product of this reaction has to be free of nitrogen. Based on chemical
reasoning, the missing part could be the epoxide glycidol and/or its hydrolysis product
glycerol because in the reaction neither oxygen nor other cofactors were required. However,
no indications for these compounds were found by IH NMR spectroscopy. By following the
degradation with high initial concentrations of TM, several signals in the IH NMR evolved
proportionally to TMA (Figure 5.4a and b). However, we were not able to attribute the
chemical shifts and integrals observed to any compound or product. The same signals were
found in IH NMR spectra of assays which were purified by filtration (removal of proteins)
and by desalting (removal of TM, TMA and buffer) and concentrated by freeze-drying
(Figure 5.4c). In IH NMR spectra of control assays done with inactivated protein solutions
these signals were absent. Therefore, these signals were attributed to the products formed,
which are obviously neither volatile nor charged.
101
Chapter 5
TM__--A---_
H20 ( --=~
(a):;.-- :;.-"-
TMAp
\ p
p
1p
I .1 ~ ~ 11 I
(b)TMA
p
p
p
p
(c)p p
PG
~
I I I I I I I I8CH) 5.0 4.5 4.0 3.5 3.0 2.5 2.0
I1.5 ppm
Figure 5.4. lH NMR spectra of untreated and treated samples deriving from enzyme assays performed
with the particulate fraction of TM-grown cells of P. putida TM I with TM as substrate. (a) lH NMR
spectrum of an untreated sample taken after 5 hours of incubation from an assay with 0.6 mg ml-1 of
protein and an initial TM-concentration of 45 mM. The assay was carried out using DzO-containing
buffer. (b) lH NMR spectrum of an untreated sample taken after 24 hours from the same experiment as
in (a). (c) lH NMR spectrum of a filtered, desalted and freeze-dried sample (dissolved in DzO) taken
after 2 hours from an assay with 0.85 mg ml-1 of protein and an initial TM-concentration of 7.5 mM.
P: Signals of potential products. G: Residual glycerol from membrane filter used for the removal of
the PP.
102
Microbial degradation of TM
To obtain more information on the properties of the product(s), enzyme assays were
performed with radioactive TM, using 14C-propyl-Iabelled TM. Aliquots taken after 2 hours
of incubation from assays with PF and from control assays with inactivated PF were run on
silicagel thin layer chromatography plates using either acetate: NaCI (1 M) or
methanol: ethyl acetate, both 1:1 (v/v), as eluents. The plates were then incubated on
radiosensitive films (Figure 5.5).
In this way the presence and formation of at least two products deriving from the propyl
group of TM was demonstrated. Samples were also freeze-dried and subsequently dissolved
in methanol before carrying out TLC. The same picture as shown in Figure 5.5 resulted,
confirming again that these products were not volatile.
Samples of the assays performed with 14C-Iabelled compounds were also extracted with
different solvents, such as ethyl acetate, diethyl ether and methyl acetate. Significant amounts
(a)
Acetone: NaCI (1 M)
1 : 1
(b)
Flow
Start
Ethylacetate : MeOH
1 : 1
Figure 5.5. Separation of educts and products from enzyme assays using the PF of P. putida TM 1 by
thin layer chromatography (silicagel coated alu-plates). (a) Acetone: NaCl (1 M) and (b) ethylacetate :
methanol, both 1:1 (v/v), were used as eluents. Samples were taken from active PF and heat
inactivated PF control assays with a protein concentration of 1.2 mg rnI-1and about 0.15 mM of 14C_
propyl-labelled TM. Lane 1 and 2: Samples of the control (1) and active (2) assay taken immediately
after the addition of 14C_propyl TM. Lane 3 - 6: Samples of the control (3, 5) and active (4, 6) assay
taken after two hours of incubation.
103
Chapter 5
of the 14C-propyl radioactivity attributed to the unidentified products were extractable with
the solvent ethyl acetate only. In control assays with inactivated PF and in enzyme assays
with 14C-methyl-Iabelled TM, less than 1 % of the radioactivity was extractable with this
solvent. This demonstrates that neither the educt nor TMA was extracted with ethyl acetate.
Overall, in assays using 14C-propyl-Iabelled TM about 11 % of the label were found in the
ethyl acetate layer and was attributed to (an) unidentified product(s). With the other solvents
used either a very low percentage of the extracted products was detected in the solvent layer
(diethyl ether), or primarily the educt and TMA were extracted by the solvent (methyl ether).
Purified and concentrated samples (filtered, desalted and freeze-dried) were analysed by
GC-MS, before and after trimethylsilyl derivatisation. In addition, the ethyl acetate phase of
the extraction procedure was injected in the GC-MS (underivatised). However, no significant
signals of a potential product were detected in any of the samples. Direct infusion into a mass
spectrometer (ESI-MS) of untreated samples was also performed, but this only provided the
educt signal. Therefore, with the exception of TMA the product(s) formed from TM by the
membrane-associated TM-Iyase remain unidentified so far.
Substrate specificity. The substrate specificity of the TM-transforming enzyme system in the
PF of P. putida TM 1 was tested for its activity with several structurally similar compounds.
These included the QAAs DM and MM, which are also used as head groups in esterquat
surfactants, as well as the related compounds L-, D-camitine, betaine and choline. All enzyme
assays were performed under the same conditions as those used for TM, i. e. with initial
substrate concentrations of 1.5 mM under aerobic conditions and without the addition of any
cofactors or cosubstrates. Judged from I H NMR spectroscopy analysis only choline was
converted in the PF of P. putida TM 1, but TMA was obviously not produced. Evaluation of
the I H NMR spectra provided two oxidation products of choline, namely betainealdehyde and
betaine (Figure 5.6). These structures were confirmed using pure betainealdehyde and betaine
standards (for details on the analysis, see chapter 4). In the PF of P. putida TM 1 cells grown
with choline as sole source of carbon, energy and nitrogen, choline was converted in the same
way. However, no TM-transforming activity was detected in such extracts.
104
Betaineo
lJ-°H/\
Betainealdehyde(hydrated form)
OHIJ-OH
/N\
\
Microbial degradation of TM
8(1H) 3.28 3.26 3.24 3.22 3.20 ppm
Figure 5.6. IH NMR spectrum with assignments of the methyl groups of choline and its degradation
products betaine and betainealdehyde. The sample for IH NMR spectroscopy was taken after 2 hours
of incubation of the PF (1.1 mg ml'l of protein) ofTM-grown P. putida TM 1 with 1.5 mM of choline.
Only the section of the CH3-group signals is shown.
Fate of the TM carbon in growing cells of P. putida TM 1. Carbon balances in batch
cultures of P. putida TM 1 with l4C-labelled TM were performed to investigate the fate of the
carbon atoms in TM. When l4C-methyl-labelled TM was used, - 25 % of the utilised TM
carbon was incorporated into biomass and - 75 % was combusted to C02. With l4C_propyl_
labelled TM, even - 92 % of the converted TM ended up in CO2 and only - 8 % were
incorporated into biomas (note: the propyl carbon in position 3 was labelled). Based on these
results and since P. putida TM 1 was not able to grow with Cl-compounds (Kaech & Egli,
2001), formaldehyde and formate utilising dehydrogenases were expected to be present in
cell-free extracts, mediating the oxidation of the methyl carbon to C02 and providing energy.
Therefore, the cell-free extracts of P. putida TM 1 were tested for activity of these widely
occurring enzymes. Indeed, high activity was found in the soluble fraction for both, NAD+
dependent formaldehyde and NAD+-dependent formate dehydrogenase, amounting to 362 ±
13 nmol min'I mg'I (7 measurements) and 130 ± 15 nmol min·I mg· l (5 measurements),
respectively. In control experiments with the CFE of acetate-grown cells, no formate
dehydrogenase activity, and only a low activity of formaldehyde dehydrogenase (30 nmol
min·I mg·I) was detected.
105
Chapter 5
DISCUSSION
All detected enzyme activities were associated with the membrane of strain TM 1 making
purification of the enzymes difficult or even impossible. Therefore, enzyme assays for the
investigations of the catabolic pathways were performed with the particulate fraction of strain
TM 1.
The initial degradation of TM is mediated by an inducible, membrane-associated lyase,
splitting TMA from TM (Figure 5.7a). This reaction was shown to be a physiological property
of P. putida TM 1 and no "in vitro" artifact even though externally provided TMA,
unexpectedly, was not used by growing P. putida TM 1 as a nitrogen source. The extent of
only 30 % of TM-conversion in enzyme assays using the particulate fraction was neither
caused by enantioselectivity of the enzyme system nor by an inhibition of the products. An
inactivation of the enzyme by a reactive (unidentified) product can also be excluded based on
the following "worst case" calculation. If the PF consisted solely of TM-active protein
(2 mg mr1) with a molecular weight of 10'000 Da (small protein) and each product molecule
inactivated one enzyme molecule, a consumption of 27 mg r 1 of TM might be reached only,
instead of a consumption of up to 1.8 g r 1 of TM as was observed.
In the literature the fission of the N-Calkyl bond was proposed by Van Ginkel (1996) as a
general strategy of microorganisms to degrade alkyl-amines and alkyl-trimethyl-ammonium
compounds. This mechanism was observed in Pseudomonads for hexadecyl-trimethyl
ammonium (Van Ginkel et aI., 1992), didecyl-dimethyl-ammonium (Nishihara et aI., 2000),
as well as for dodecyl-dimethyl-amine (Kroon et aI., 1994). However, the fission of the N
Calkyl bond was always catalysed by an oxidation reaction and the degradation of these
compounds was either restricted to a consortium of at least two microorganisms or was
incomplete resulting in the excretion of the corresponding amine into the culture medium. The
ability to split TMA from choline, different betaines and/or carnitines has also been reported
for various microorganisms and, where investigated, the fission was found to be mediated by
an oxidation reaction (Englard et aI., 1983; Kleber, 1997; Rebouche & Seim, 1998; Seim et
aI., 1982a; Seim et aI., 1982b; Unemoto et aI., 1966; Wood & Keeping, 1944). In contrast to
these reports, none of these quaternary trimethyl-ammonium compounds (i. e. choline, betaine
or L, D-camitine) was split to TMA and corresponding products in the PF of P. putida TM 1.
The N-Cethanol bonds of DM and MM were not split by the PF of TM-grown P. putida TM 1
106
Microbial degradation of TM
either. Therefore, the lyase activity in P. putida TM 1 must be very specific and the presence
of a quaternary trimethyl-ammonium center in substrates is obviously not sufficient enough to
allow fission of the N-Calkyl bond.
(a)OH
lpOH I[ ? ]
I ~OH/\ •
/N"",+
oGlycidolTM Trimethylamine
: .,.H2OI "
? l0 / [HO~OH]
H)lHHN
'" Gly::rolFormaldehyde Dimethylamine
: .,.ATPI "
~n ? l2 e-+ 2 H+
0 t 0 JH)lOHH
2N- \\/OH
PHO~O/\
Formate MethylamineG~~erol-3P OH
~mI
? III
2 e-+ 2 H+ •CO2 NH3
Central metabolism
(b)
2 e-+ 2 H+ HO
l0°~. IJ-OH/ \ -/~ /N\
Choline H20 Betainealdehyde(hydrated form)
Betaine
Figure 5.7. Ca) Proposed degradation pathway of TM: I) "TM-Iyase", II) NAD+-dependent
formaldehyde dehydrogenase, Ill) NAD+-dependent formate dehydrogenase. Cb) Oxidation of choline
in the particulate fraction of TM-grown P. putida TM 1.
107
Chapter 5
If the first step in TM catabolism in cells of P. putida TM 1 leads to the production of TMA,
and this metabolite is not excreted into the medium but used as a source of nitrogen by the
bacterium, one must expect TMA to be demethylated in some ways (Figure 5.7a). However,
neither TMA dehydrogenase nor monooxygenase activity was found employing the classic
assays described in the literature for methylotrophs (Colby & Zatman, 1972). The inability to
detect activity does not exclude the involvement in particular of TMA monooxygenase in
P. putida TM 1 because these enzymes are known to be very labile (Boulton et aI., 1974). The
removal of methyl groups from TMA is indicated by the high activity of inducible, NAD+
dependent formaldehyde and formate dehydrogenase in the soluble protein fraction,
suggesting that the methyl groups were oxidised to CO2 rather than incorporated into biomass
(Figure 5.7a). This suggestion is supported by the observation that P. putida TM 1 was unable
to grow with Cl-compounds (Kaech & Egli, 2001). Carbon balances with 14C-methyl-Iabelled
TM confirmed the complete oxidation of the methyl groups, since 75 % of the methyl carbon
was transformed to C02. Astonishingly, not only the Cl groups but also the 14C_propyl_
labelled carbon was almost exclusively combusted to CO2, this carbon atom of the molecule
even to an extent of 92 %. These results correspond with the low growth yield recorded in
batch cultures of P. putida TM 1 and the high percentage of 60 % of the total carbon present
in this compound combusted to CO2 (Kaech & Egli, 2001).
Considerable effort was put into the identification of the remaining part of the TM molecule.
The formation of at least two products, neither charged nor volatile, was demonstrated but
elucidation of their structures by performing GC-MS, ESI-MS and IH NMR was not
successful. Future investigations should focus on separating and concentrating these
compounds. Best chances for success will probably involve the development of an appropriate
liquid chromatography method, coupled to mass spectrometry.
With respect to the substrate specificity of the TM-transforming enzyme system in the PF of
P. putida TM 1 grown with TM, only choline was converted, whereas DM, MM, betaine and
carnitine remained untouched. However, when compared to TM, the conversion of choline
was achieved by a totally different mechanism. Choline was oxidised to betainealdehyde and
betaine in the PF of P. putida TM 1 indicating the presence of a membrane-associated choline
oxidoreductase mediating both oxidation steps (Figure 5.7b). The same activity was found in
the PF of isolate MM 1, too (Chapter 4). The oxidation of choline to betainealdehyde and
108
Microbial degradation of TM
betaine was reported in the literature to be the most frequently occurring pathway for choline
degradation in bacteria (Kortstee, 1970; Shieh, 1964). Several membrane-associated, choline
specific enzymes are documented in the literature for Pseudomonas aeruginosa and
Escherichia choli (Bater & Venables, 1977; Lamark et aI., 1991; Nagasawa et al., 1976;
Russell & Scopes, 1994). Interestingly, all these choline oxidoreductases (dehydrogenases)
only oxidised choline to betainealdehyde, but did not catalyse a second oxidation step to the
acid (betaine) as found in this study. Since two completely different mechanisms in the PF
were responsible for the initial degradation of choline and TM and because TM was not
converted in the PF of choline-grown cells, the enzymes acting on choline and TM are
obviously distinctly different.
ACKNOWLEDGEMENTS
We thank Hans-Ruedi Aemi and Rene Schonenberger for the GC-MS and ESI-MS analysis.
109
ite Leer /Blank leaf
Concluding remarks
6. Concluding remarks
Detailed discussion of the results can be found in the individual chapters. Here, I would like
to point out and add some conclusions and hypotheses related to the questions addressed at
the beginning of this work.
Competent bacteria. With all three quaternary ammonium alcohols TM, DM and MM,
microorganisms were isolated that were able to grow with one or more of these QAAs as a
sole source of carbon, energy and nitrogen. So far, the degradation of quaternary ammonium
compounds was supposed to be achieved exclusively by consortia of at least two different
microorganisms (Van Ginkel, 1996). However, this hypothesis was based on investigations of
mainly long alkyl chain quaternary ammonium compounds whereas the QAAs investigated in
this study contained short hydroxylated alkyl substituents and the properties of these short
substituents may be the key to this issue. Anyhow, the general view that consortia are
required for complete degradation of quaternary ammonium compounds must be questioned.
In contrast to earlier reports the QAAs were provided here as the sole source of carbon,
energy and nitrogen and - probably because of this enrichment strategy - the isolated
microorganisms were able to completely catabolise the individual QAAs. Nevertheless, in
batch cultures not each of the isolated strains degraded all the QAA supplied to completion,
yielding exclusively biomass, NH/ and CO2. Strain MM 1, when cultivated with MM,
excreted considerable amounts of dead-end metabolites (Figure 6.2), and strain DM 1 used
only a part of the DM supplied (why is not known, tests done for limitiations in the growth
medium indicated excess of all essential nutrients) and, therefore, these compounds have the
potential to accumulate in the environment. These findings could have considerable impact on
the design of new similar compounds with respect to enhanced environmental properties, i. e.,
ready and ultimate biodegradation.
Interestingly, standard biodegradation tests with activated sludge or waste water treatment
effluent (Hales, 1998; Krueger et aI., 1998; Puchta et al., 1993; Waters et al., 1991) suggested
degradation to completion for all three QAAs, TM, DM and MM, forming biomass and CO2
of the carbon provided. Complete degradation in complex systems could be to due not only to
strains similar to these isolated here able to degrade QAAs intracellularly to completion, but
to simultaneous consumption of additional carbon sources to balance the nitrogen excess of
QAAs (see Egli, 1995), or additional microorganisms degrading metabolites from QAAs
excreted by "incomplete degraders". This raises the question, whether or not the microbes we
111
Chapter 6
have isolated and characterised are the same as those that do the job in the environment. To
answer this question, however, further investigations are required. Adequate methods to
investigate the abundance of the isolated bacterial strains in environmental systems exposed
to QAAs would be the use of 16S-rRNA probes for fluorescent in situ hybridisation, since
16S-rDNA was sequenced for all isolated strains, or surface antibodies like used for NTA
degrading bacteria (Bally et aI., 1994; Wilberg et aI., 1993).
The four isolated characterised strains differ considerably with respect to their physiological,
nutritional and biochemical properties. This supports and confirms the findings for the
degradation of the three QAAs in standardised biodegradation tests reported by Hales (1998),
who observed different degradation patterns. Phylogenetically, i. e. based on 16S-rDNA
similarity, the isolated bacteria belonged not only to different microbial genera but the isolates
are positioned in each of the major bacterial groups of protobacteria (Figure 6.1). The ability
to degrade TM, DM or MM is obviously not a trait of a single bacterial genus or even a
particular species.
Surprisingly, only one of the isolated strains, named DM 1, was able to grow with another
QAA (TM) than that used for its isolation, whereas all of them were able to degrade the
natural, structurally related compound choline. Vice versa, none of the tested established
reference choline degraders P. putida DSM 291T and Z ramigera Itzigsohn 1868AL or several
bacterial strains isolated in the course of this work with choline were able to degrade any of
the QAAs. Hence, the ability to degrade TM, DM or MM appears to be a specific property of
microorganisms and the competence to degrade choline does not necessarily imply the ability
to degrade one of the QAAs derived from esterquat surfactants. Therefore, the design of
QAAs with a structure similar to that of choline is not a priori a prerequisite for favorable
environmental properties.
112
Concluding remarks
Proteobacteria
a Subdivision
AcetobacteraceaeCau/obacter groupHyphomonas groupRhizobiaceae groupRhodobactergroup ------------------------------------ Isolate MM 1 (new genus?)Rhodospirillaceae MM-oxidoreductaseRickettsia/es (choline oxidoreductase?)SphingomonadaceaeUnclassified alpha proteobacteria
~ Subdivision
A/ca/igenaceae iAmmonia-oxidizing bacteria :--Burkho/deria/Oxa/obacter/Ra/stonia group El
Comamonadaceae Isolate OM 1 (new species?)Gallionella group , Isolate OM 2 (new species?)Hydrogenophilus group I I..-
Massilia El l--Methy/ophilus group J= Zoog/oea ramigeraNeisseriaceae Zoog/oea resiniphilaRhodocyc/us group --------'. Zoog/oea Zoog/oea sp.Spirillum groupUnclassified beta proteobacteria
y Subdivision
AeromonadaceaelSuccinivibrionaceae groupAlishewanella groupA/teromonadaceae groupCardiobacteriaceae groupChromatiaceaelEctothiorhodospiraceae groupCuracaobacter groupEnterobacteriaceae groupLegionellaceae groupMethy/ococcaceae group ~OceanospirillumlHa/omonas groupPasteurellaceae groupPseudomonaceaelMoraxellaceae group P. putida -- P. putida TM 1SUlfur-oxidizing symbionts TM-IyaseThiothrixlFrancisella groupVibrionaceae groupXanthomonada/esUnclassified gamma proteobacteria
o/E Subdivision
Unclassified proteobacteria
Figure 6.1. Phylogenetic tree of the proteobacteria according to the National Center for Biotechnology
Information (NCBI, www.ncbi.nlm.nih.gov)andpositionsoftheisolatesTM1.DM1.DM 2 and
MM 1.
113
Chapter 6
Initial degradation steps. Consumption of MM and TM was observed in the cell-free
extracts of the strains MM 1 and TM 1, respectively. Figure 6.2 gives an overview over the
initial degradation steps of MM, TM and choline in the microbial isolates based on the
observations in this work. Completely different enzymatic mechanisms were responsible for
the degradation of these two QAAs confirming again the different degradation patterns found
for the QAAs in standardised biodegradation tests (Hales, 1998).
The initial degradation of MM in strain MM 1 was mediated by a membrane-associated,
constitutively expressed oxidoreductase. Since also DM and choline were oxidised exhibiting
the same characteristics, the responsible enzyme for this reaction was probably the same
enzyme. Two observations suggest that this enzyme may be a choline oxidoreductase with
extended substrate specificity. First, its ability to degrade choline, second that it is
constitutively expressed during growth with choline, MM or acetate. Hence, the initial
degradation of MM is most probably linked to the degradation of choline. Investigation of this
relationship could be performed by producing choline-deficient mutants of strain MM 1 and
checking them for the ability to degrade MM. DNA sequences of several genes encoding
choline dehydrogenases and betainealdehyde dehydrogenases have been analysed (Lamark et
al., 1991; Pocard et al., 1997; Rosenstein et al., 1999), and could be a basis for investigations
at the genetic level.
The fact that TM was not transformed in the cell-free extracts of strain MM 1 can be
interpreted in two ways: Either, the presence of an ethanol group in the substrate is essential
to undergo oxidation in the enzyme and neither the presence of hydroxyl groups alone nor the
quaternary ammonium character of the compound is sufficient to evoke enzymatic oxidation
of hydroxyl groups; or the 2,3-dihydroxy-propyl part of TM, i. e. its steric dimension (Figure
6.2), prevents the compound to fit to the active site of the protein.
The initial degradation of TM in strain TM 1 followed a completely different mechanism.
Trimethylamine was split from TM by an inducible, membrane-associated lyase. None of the
structurally similar compounds DM, MM, choline, betaine or carnitine were accepted as
substrates by this enzyme system. However, choline underwent oxidation in the particulate
fraction of strain TM 1 providing betainealdehyde and betaine, as was also observed in strain
MM 1. Therefore, the trimethyl-ammonium structure is not sufficient to have a QAA accepted
as a substrate, suggesting that the lyase reaction is very specific. The same mechanism for the
initial TM-fission also was found in strain DM 2 isolated with DM, which was also growing
with TM. This strain belonged to a different genus, which leads to the speculation, whether or
not the degradation of TM is mediated predominantly or even generally by the mechanism
114
Concluding remarks
described above. However, to answer this question extensive screening of TM degrading
strains would be needed.
Interestingly, the enzyme(s) oxidising choline to betainealdehyde and subsequently to betaine
in both strains, MM 1 and TM 1, were membrane-associated. In contrast, either choline
oxidising enzymes found in microorganisms catalysing both oxidation steps were reported to
be soluble (Ikuta et aI., 1977; Ohta-Fukuyama et al., 1980) or a membrane-associated enzyme
was detected mediating exclusively the oxidation of choline to betainealdehyde (Bater &
Venables, 1977; Lamark et aI., 1991; Nagasawa et aI., 1976; Russell & Scopes, 1994).
In the cell-free extracts of the DM-growing strains, no transformation of DM was detected
and, hence, making investigations at the enzyme level impossible. The reasons may be due to
inactivation of the enzymes by the preparation procedure or the storage.
Obviously, for the initial degradation of TM and MM in bacteria, no general strategy (Van
Ginkel, 1995) can be proposed and no general relationship exists between (initial) QAA and
choline degradation. The mechanisms seem to depend on the specific, individual structure of
the quaternary ammonium alcohols TM, DM and MM. Although several indications were
found with respect to the environmental properties of the QAAs TM, DM and MM, no
general conclusions can yet be drawn concerning the design of new QAAs for cationic
detergents. Each new QAA requires extensive investigations of its biodegradability and of the
potential to generate dead-end metabolites that are excreted and perhaps recalcitrant.
115
Chapter 6
(a) (b)
TM
?
I [~OH]/N".... + 0
Trimethylamine Glycidol
?~OHN )l
'" H HFormaldehyde
}...NADHo
H)lOHFormate
}...NADHCO2
MM
HO~ +r-J0H
/N~OH
°HO~ OH
N+""""~/ v--- --. Excretion asdead-end metabolite
~2e+2H.
~.,.j~OH~/ v---0
~H20 Further
2 e- + 2 H+ breakdown
OH /
HO~ OH/
N+""""!\/ v---0
~2e-+2H.
(c)
~00H---L:+ IJ-OH/ \ . IIIr-~ /N\
Choline H20 Betainealdehyde(hydrated form)
2 e-+ 2 H+ HO.L IJo
~ /N\III /'
Betaine
Figure 6.2. (a) Initial degradation steps of MM in strain MM 1. I) Constitutively expressed,
membrane-associated MM-oxidoreductase. (b) Initial degradation step of TM in P. putida TM 1 and
proposed breakdown of trimethylamine. 11) Inducible, membrane-associated TM-Iyase. (c) Observed
oxidation of choline in strain MM 1 and P. putida TM 1. Ill) Membrane-associated oxidoreductase(s).
116
References
References
Bally, M., Wilberg, E., Kuhni, M. & Egli, T. (1994). Growth and regulation of enzyme
synthesis in the nitrilotriacetic acid (NTA)-degrading bacterium Chelatobacter heintzii ATCC
29600. Microbiology 140, 1927-1936.
Bater, A. J. & Venables, W. A. (1977). The characterization of inducible dehydrogenases
specific for the oxidation of D-alanine, allohydroxy-D-proline, choline and sarcosine as
peripheral membrane proteins in Pseudomonas aeruginosa. Biochim Biophys Acta 468, 209
226.
Bax, A. (1985). A spatially selective composite 90 degrees radiofrequency pulse. J Magn
Reson 65, 142-145.
Bax, A., Griffey, R. H. & Hawkins, B. L. (1983). Correlation of proton and N-15 chemical
shifts by multiple quantum NMR. J Magn Reson 55, 301-315.
Berenbold, H. (1990). Nutzen und Bioverhalten von Weichsptilem. Tenside SurfDet 27, 34
40.
Boncompagni, E., Osterhas, M., Poggi, M.-C. & Le Rudulier, D. (1999). Occurrence of
choline and glycine betaine uptake and metabolism in the family Rhizobiaceae and their role
in osmoprotection. Appl Environ Microbiol 65, 2072-2077.
Boulton, C. A., Crabbe, M. J. C. & Large, P. J. (1974). Partial purification of a
trimethylamine mono-oxygenase from Pseudomonas aminovorans and its role in growth on
trimethylamine. Biochem J 140, 253-263.
Buck, J. D. (1982). Nonstaining (KOH) method for determination of Gram reactions of
marine bacteria. Appl Environ Microbiol44, 992-993.
Callely, A. G., Forster, C. F. & Stafford, D. A. (1977). Treatment of industrial e.fJluents.
London: Hodder & Stoughton.
Colby, J. & Zatman, L. J. (1972). Trimethylamine metabolism in obligate and facultative
methylotrophs. J Biochem (Tokyo) 132, 101-112.
117
References
Davis, A. L., Keeler, J., Laue, E. D. & Moskau, D. (1992). Experiments for recording pure
absorption heteronuclear correlation spectra using pulsed field gradients. J Magn Reson 98,
207-216.
de Vos, P. & de Ley, J. (1983). Intra- and intergeneric similarities of Pseudomonas and
Xanthomonas ribosomal ribonucleic acid cistrons. 1nt J Syst Bacteriol33, 487-509.
Dean-Raymond, D. & Alexander, M. (1977). Bacterial metabolism of quaternary
ammonium compounds. Appl Environ Microbiol33, 1037-1041.
Dess, D. B. & Martin, J. C. (1983). Readily accessible 12-1-5 oxidant for the conversion of
primary and secondary a1cohols to aldehydes and ketones. J Org Chem 48, 4155-4156.
Drews, G. (1968). Mikrobiologisches Praktikum fuer Naturwissenschaftler. Berlin: Springer
Verlag.
Dworkin, M., Schleifer, K.-H. & Stackebrandt, E. (1999-2002). The Prokaryotes. An
Evolving Electronic Database for the Microbiological Community. New York: Springer
Verlag New York, Inc.
Egli, T. (1995). The ecological and physiological significance of growth of heterotrophic
microorganisms with mixtures of substrates. Adv Microb Eco114, 305-386.
Egli, T. (2000). Nutrition of microorganisms. In Encyclopedia ofMicrobiology, pp. 431-449.
Edited by J. Lederberg, M. Alexander, B. R. Bloom, D. Hopwood, R. Hull, B. H. Iglewiski,
A. I. Laskin, S. G. Oliver, M. Schaechter & W. C. Summers. London: Academic Press.
Englard, S., Blanchard, J. S. & Miura-Fraboni, J. (1983). Production of trimethylamine
from structurally related trimethylammonium compounds by resting cell suspensions of y
butyrobetain- and D,L-camitine-grown Acinetobacter calcoaceticus and Pseudomonas putida.
Arch Microbiol135, 305-310.
Friedman, B. A. & Dugan, P. R. (1968). Identification of Zoogloea species and the
relationship to zoogloeal matrix and floc formation. J Bacteriol95, 1903-1909.
118
References
Friedman, B. A., Dugan, P. R, Pfister, R. M. & Remsen, C. C. (1969). Structure of
exocellular polymers and their relationship to bacterial flocculation. J Bacteriol 98, 1328
1334.
Garcia, M. T., Campos, E., Sanchez-Leal, J. & Ribosa, I. (2000). Anaerobic degradation
and toxicity of commercial cationic surfactants in anaerobic screening tests. Chemosphere 41,
705-710.
Garcia, M. T., Ribosa, I., Guindulain, T., Sanchez-Leal, J. & Vives-Rego, J. (2001). Fate
and effect of monoalkyl quaternary ammonium surfactants in the aquatic environment.
Environ Pollut 111,169-175.
Giolando, S. T., Rapaport, RA., Larson, R. J., Federle, T. W., Stalmans, M. &
Masscheleyn, P. (1995). Environmental fate and effects of DEEDMAC: A new rapidly
biodegradable cationic surfactant for use in fabric softeners. Chemosphere 30, 1067-1083.
Gish, W. (1996-1999). http://blast.wustl.edu.
Hales, S. G. (1998). Unpublished data.
Haubrich, D. R & Gerber, N. H. (1981). Choline dehydrogenase. Biochem Pharmacol30,
2993-3000.
Hellberg, P. E., Bergstroem, K. & Holmberg, K. (2000). Cleavable surfactants. J SurfDet
3,81-91.
Ikuta, S., Matuura, K., Imamura, S., Misaki, H. & Horiuti, Y. (1977). Oxidative pathway
of choline to betaine in the soluble fraction prepared from Arthrobacter globiformis. J
Biochem (Tokyo) 82, 157-163.
Jeener, J., Meier, B. H., Bachman, P. & Ernst, R. R (1979). Investigation of exchange
processes by two-dimensional NMR spectroscopy. J Chem Phys 71, 4546-4553.
Kaech, A. & Egli, T. (2001). Isolation and characterization of a Pseudomonas putida strain
able to grow with trimethyl-1,2-dihydroxy-propyl-ammonium as sole source of carbon,
energy and nitrogen. Syst Appl Microbiol24, 252-261.
119
References
Kiene, R. P. (1998). Uptake of choline and its conversion to glycine betaine by bacteria in
estuarine waters. Appl Environ Microbiol 64, 1045-1051.
Kleber, H. P. (1997). Bacterial carnitine metabolism. FEMS Microbiol Left 147, 1-9.
Koolman, J. & Rijhm, K. H. (1998). Taschenatlas der Biochemie, second edn. Stuttgart,
New York: Georg Thieme Verlag.
Kortstee, G. J. J. (1970). The aerobic decomposition of choline by microorganisms. Arch
Microbiol71, 235-244.
Kroon, A. G. M., Pomper, M. A. & van Ginkel, C. G. (1994). Metabolism of
dodecyldimethylamine by Pseudomonas MA3. Appl Microbiol Biotechnol42, 134-139.
Krueger, G., Boltersdorf, D. & Overkempe, K. (1998). Esterquats. In Novel surfactants,
pp. 116-139. Edited by K. Holmberg. New York: Marcel Dekker.
Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of
bacteriophage T4. Nature 227,680-685.
Lamark, D., Kaasen, I., Eshoo, M. W., Falkenberg, P., McDougall, J. & Strom, A. R.
(1991). DNA sequence and analysis of the bet genes encoding the osmoregulatory choline
glycine betaine pathway of Escherichia coli. Mol Microbiol5, 1049-1064.
Matthijs, E., Debaere, G., Itrich, N., Masscheleyn, P., Rottiers, A., Stalmans, M. &
Federle, T. (1995). The fate of detergent surfactants in sewer systems. Water Sci Technol31,
321-328.
Nagasawa, T., Kawabata, Y., Tani, Y. & Ogata, K. (1975). Choline dehydrogenase of
Pseudomonas aeruginosa A-16. Agric Bioi Chem 39, 1513-1514.
Nagasawa, T., Mori, N., Tani, Y. & Ogata, K. (1976). Characerization of choline
dehydrogenase from Pseudomonas aeruginosa A-16. Agric Bioi Chem 40, 2077-2084.
120
References
Nishihara, T., Okamoto, T. & Nishiyama, N. (2000). Biodegradation of
didecyldimethylammonium chloride by Pseudomonas jluorescens TN4 isolated from
activated sludge. Appl Environ Microbiol88, 641-647.
OECD (1981). OECD Guidelines for the testing of chemicals, Section 3, Degradation and
accumulation. Paris: Organisation for Economic Cooperation and Development (OECD).
Ogawa, A. & Curran, D. P. (1997). Benzotrifluorid: A useful alternative solvent for organic
reactions currently conducted in dichloromethane and related solvents. J Org Chem 62, 450
451.
Ohta-Fukuyama, M., Miyake, Y., Emi, S. & Yamano, T. (1980). Identification and
properties of the prosthetic group of choline oxidase from Alcaligenes sp. J Biochem (Tokyo)
88, 197-203.
Palmer, A. G., Cavanagh, J., Wright, P. E. & Ranee, M. (1991). Sensitivity improvement
in proton-detected 2-dimensional heteronuclear correlation NMR-spectroscopy. J Magn
Reson 93,515-170.
Pfennig, N., Widdel, F. & Trueper, H. G. (1981). The dissimilatory sulfate-reducing
bacteria. In The Prokaryotes, pp. 931. Edited by M. P. Starr, H. Stolp, H. G. Trueper, A.
Balows & H. G. Schlegel. Berlin: Springer Verlag.
Pocard, J.-A., Vincent, N., Boncompagni, E., Smith, L. T., Poggi, M.-C. & Le Rudulier,
D. (1997). Molecular characterization of the bet genes encoding glycine betaine sysnthesis in
Sinorhizobium meliloti 102F34. Microbiology 143, 1369-1379.
Puchta, R., Krings, P. & Sandkiihler, P. (1993). A new generation of softeners. Tenside
SurfDet 30, 186-191.
Rebouche, C. J. & Seim, H. (1998). Carnitine metabolism and its regulation in
microorganisms and mammals. Annu Rev Nutr 18, 39-61.
121
References
Rosenstein, R., Futter-Bryniok, D. & Goetz, F. (1999). The choline-converting pathway in
Staphylococcus xylosus C2A: Genetic and physiological characterization. J Bacteriol 181,
2273-2278.
Russell, R. & Scopes, R. K. (1994). Use of hydrophobic chromatography for purification of
the membrane-located choline dehydrogenase from a Pseudomonas strain. Bioseparation 4,
279-284.
Sambrook, J., Frisch, E. F. & Maniatis, T. (1989). Molecular cloning: a laboratory manual.
Cold Spring Harbor: Cold Spring Harbor Laboratory Press.
Scheiner, D. (1976). Determination of ammonia and Kjedahl nitrogen by indophenol method.
Water Res 10, 31-36.
Seim, H., Claus, R., Kleber, H. P. & Strack, E. (1982a). Splitting of the C-N bond in
carnitine by an enzyme (trimethyl-forming) from membranes of Acinetobacter calcoaceticus.
FEMS Microbiol Left 15, 165-167.
Seim, H., Loster, H., Claus, R., Kleber, H. P. & Strack, E. (1982b). Formation of y
butyrobetaine and trimethylamine from quaternary ammonium compounds structure-related to
L-carnitnine and choline by Proteus vulgars. FEMS Microbiol Left 13, 201-205.
Shaka, A. J., Barker, P. B. & Freeman, R. (1985). Computer-optimized decoupling scheme
for wideband applications and low-level operation. J Magn Reson 64, 547-552.
Shaka, A. J., Keeler, J., Frenkiel, T. & Freeman, R. (1983). An improved sequence for
broadband decoupling: WALTZ-16. J Magn Reson 52,335-338.
Shen, M.-K. (1988). Spectrophotometric determination of trimethylamine nitrogen in raw
waters using picric acid. Analyst 113, 1139-1140.
Shieh, H. S. (1964). Aerobic degradation of choline. Can J Microbiol10, 837-842.
Simms, J. R., Woods, D. A., Walley, D. R., Keough, T., Schwab, B. S. & Larson, R. J.
(1992). Integrated approach to surfactant environmental safety assessment: Fast atom
122
References
bombardment mass spectrometry and liquid scintillation counting to determine the
mechanism and kinetics of surfactant biodegradation. Anal Chem 64, 2951-2957.
Sullivan, D. E. (1982). Biodegradation of a cationic surfactant in activated sludge. Water Res
17, 1145-1151.
Unemoto, T., Hayashi, M., Miyaki, K. & Hayashi, M. (1966). Formation of trimethylamine
from DL-carnitine by Serratia marcescens. Biochim Biophys Acta 121, 217-220.
Unz, R. F. (1971). Neotype strain of Zoogloea ramigera Itzigsohn. Int J Syst Bacteriol 21,
91-99.
Valles, B., Riva, M. c., Sappera, C., Canela, M. & Griful, E. (2000). Toxicity values for
cationic surfactants. Tenside SurfDet 37, 290-295.
van Ginkel, C. G. (1995). Biodegradability of cationic surfactants. In Biodegradability of
surfactants, pp. 183-203. Edited by D. R. Karsa & M. R. Porter. London: Blackie academic
and professional.
van Ginkel, C. G. (1996). Complete degradation of xenobiotic surfactants by consortia of
aerobic microorganisms. Biodegradation 7, 151-164.
van Ginkel, C. G., Dijk, J. B. & Kroon, A. G. M. (1992). Metabolism of
hexadecyltrimethylammonium chloride in Pseudomonas strain B1. Appl Environ Microbiol
58, 3083-3087.
Vievsky, A. (1997). Cationic surfactants: new perspectives in medicine and biology. Tenside
SurfDet 34, 18-21.
Waters, J., Kleiser, H. H., How, M. J., Barratt, M. D., Birch, R. R., Fletcher, R. J.,
Haigh, S. D., Hales, S. G., Marshall, S. J. & Pestell, T. C. (1991). A new rinse conditioner
active with improved environmental properties. Tenside SurfDet 28, 460-469.
Waters, J., Lee, K. S., Perchard, V., Flanagan, M. & Clarke, P. (2000). Monitoring of
diester cationic surfactant residues in UK and Dutch sewage treatment effluents. Tenside Surf
Det 37, 161-171.
123